garrytan/gstack
40.7
Weak · 14 November 2026
195.7k
lines of production code
TypeScript
primary language
7
measurements over time
What this system is
Features
Added Hacker News frontpage skill with browser automation client
The browser-skills/hackernews-frontpage location now includes a new browse-client SDK (browse-client.ts) that allows the skill to drive a browser via a local gstack daemon. This client handles authentication through environment variables or a state file, manages timeouts, and provides methods for navigation and interaction, enabling the Hacker News frontpage skill to automate browser tasks.
browser-skills/hackernews-frontpage · high confidence
Added UX principles doctrine for design decisions
A new 'doctrine' section has been added to the design-shotgun knowledge base, establishing core UX principles such as minimizing cognitive load, designing for scanning, and prioritizing clarity over consistency. This content is now registered in the section manifest and is automatically triggered when users are writing variant concepts or design briefs, ensuring these behavioral guidelines inform the design direction.
design-shotgun/sections · high confidence
Added Windows Node.js server build script and extension ID utility
This change introduces two new scripts in the browse/scripts directory to support Windows environments and extension security. The build-node-server.sh script generates a Node.js-compatible server bundle for Windows, addressing Playwright compatibility issues on that platform by transpiling the server code and injecting Bun API polyfills. Additionally, the extension-id.ts script provides a utility to derive the Chrome extension ID from the public key in the manifest, enabling the browse server to verify the Origin header against a known extension identity for secure token handling.
browse/scripts · high confidence
Added contributor documentation for creating new host configurations
A new template file (SKILL.md.tmpl) has been added to the contrib/add-host directory to guide contributors in adding support for new AI coding agents. This document outlines the steps to create a host configuration file, including gathering host-specific information, creating the TypeScript config, registering it in the index, updating .gitignore, and verifying the setup through documentation generation and tests.
contrib · high confidence
Added exploratory QA and reporting sections to the qa-only skill
The qa-only skill now includes structured documentation for report-only exploratory QA and evidence-grounded report finalization. These new sections define the workflow for conducting probes (with specific timing budgets and evidence recording tools), classifying findings, and assembling final reports based strictly on retained observations, ensuring that claims are grounded in actual command results rather than speculation.
qa-only/sections · high confidence
Added shared-code extraction audit skill
The deslop-shared-libs area now includes a new SKILL.md and its template (SKILL.md.tmpl) that define a read-only audit process for identifying opportunities to extract shared code. This skill guides the user to review recent commits and open pull requests, evaluate candidates against a specific rubric (proving callers, reuse before extraction, keeping helpers small), and output ranked recommendations for the best three opportunities, while strictly enforcing safety boundaries such as using a safe Git wrapper and avoiding file writes or script execution.
deslop-shared-libs · high confidence
Added shell script helpers for binary discovery and remote identification
The browse/bin directory now includes two new executable scripts: find-browse, which acts as a shim to delegate to a compiled binary or fall back to discovering the browse tool via standard agent skill directories, and remote-slug, which extracts the owner-repo identifier from the git remote URL to support project-specific path resolution.
browse/bin · high confidence
CEO review skill adds Selective Expansion mode and cross-session context memory
The plan-ceo-review skill now includes a fourth review mode, Selective Expansion, allowing users to cherry-pick specific scope expansions while holding the rest of the plan. The skill also gains cross-session memory via gbrain context queries, automatically retrieving prior CEO plans, recent design docs, and past review activity to provide continuous, informed reviews. Additionally, it now supports WebSearch for landscape checks and includes a handoff mechanism to resume paused reviews from /office-hours sessions.
plan-ceo-review · high confidence
CSO runtime and scanner qualification infrastructure
The CSO library now includes a complete, hardened runtime and scanner qualification system. This introduces a new CLI (gstack-cso v3.0.0) and a set of new modules that manage the lifecycle of immutable, SHA-256-digested Docker images for Node, Bun, Python, and Rails. The system enforces strict security boundaries through native C launchers (launcher.c, launcher-windows.c) that scrub environment variables and verify the core binary's hash before execution. It also adds a bounded, race-safe file system cache, a resource-admission controller for Docker containers, and a reproducible recovery mechanism for failed verification attempts, ensuring that runtime staging and scanner catalogs can be shipped with verified integrity.
lib/cso · high confidence
Codex skill adds Challenge and Consult modes with improved error handling
The Codex skill now supports two new interaction modes alongside the existing Review mode: Challenge mode (Step 2B) for adversarial code analysis and Consult mode (Step 2C) for free-form questions with session continuity. These modes are defined in new \challenge-mode.md\ and \consult-mode.md\ files (generated from templates) and registered in the \manifest.json\. The update also introduces robust error handling across all modes: Codex now captures and surfaces authentication errors and non-zero exit codes instead of failing silently, and it detects and reports stalls (timeouts) with actionable messages to prevent misdiagnosis of model API issues.
codex/sections · high confidence
Declarative multi-host platform with new host integrations
The system now supports a declarative configuration model for external AI coding assistants, replacing ad-hoc setups with a unified \defineHost\ factory that standardizes path rewrites, tool-name mappings, and capability profiles. This change introduces built-in support for GitHub Copilot CLI, Google Antigravity CLI (agy), Cursor, Kiro, OpenCode, Slate, OpenClaw, Hermes, and GBrain, each with specific tool glossaries and skill path mappings. The primary host remains Claude Code, while external hosts are configured with appropriate safety hooks, plan modes, and delegation capabilities, allowing users to run the same skill suite across different AI providers with consistent behavior.
hosts · high confidence
Initial CSO scanner image qualification framework and catalog
This change introduces the foundational infrastructure for the CSO scanner image system, including a reviewed build-inputs manifest, a qualification workflow definition, and a wrapper Dockerfile that standardizes scanner images to a fixed non-root user and entrypoint. It also adds the initial qualified scanner catalog containing images for Gitleaks, OSV, Semgrep, Zizmor, Trivy, and Schemathesis, along with documentation detailing the promotion and verification gates required to update the catalog.
lib/cso/scanner-images · high confidence
Introduce 'careful' skill for destructive command guardrails
A new 'careful' skill has been added to provide safety guardrails for destructive commands. When activated (via triggers like 'be careful' or 'safety mode'), it intercepts Bash commands to warn users before executing high-risk operations such as \rm -rf\, \DROP TABLE\, force-pushes, or \kubectl delete\. The skill distinguishes between MEDIUM warnings (which allow user override) and HIGH tier denials (hard stops for catastrophic actions like deleting the home directory or force-pushing to the default branch). It also supports additive, project-specific custom patterns via configuration files.
careful · high confidence
Introduce 'learn' skill for managing project knowledge
A new 'learn' skill has been added to the system, allowing users to review, search, prune, and export project learnings captured across sessions. The skill provides commands to show recent insights, search by query, prune stale or contradictory entries, export knowledge to markdown, and view statistics. It integrates with the gstack infrastructure via a preamble script and supports both interactive and spawned session modes, including specific handling for AskUserQuestion failures and plan mode operations.
learn · high confidence
Introduce /autoplan publication guard with bounded journal reading and audit logging
The /autoplan workflow now enforces a strict publication guard that validates phase entries against the Claude Code session journal before allowing them. This guard reads a bounded, content-free index of the session journal to verify phase identity, snapshot integrity, and correct sequencing, ensuring that only authorized and correctly formed phase reports are published. It includes a logging mechanism that records allow/deny decisions and dispositions for audit purposes, and provides detailed, actionable error messages for various failure modes such as missing snapshots, incorrect phase ordering, or unauthorized modifications. This change enhances the reliability and security of the /autoplan process by preventing unauthorized or malformed phase entries from being published. Also in this area: Updated proactive skill defaults and Windows shell compatibility; Autoplan now correctly detects linked worktrees when the project directory is nested below the repository root.
autoplan/bin · high confidence
Introduce /qa skill for systematic web application testing and bug fixing
The new /qa skill enables the agent to systematically test web applications and iteratively fix bugs found during testing. It supports three testing tiers (Quick, Standard, Exhaustive) to balance coverage and speed, and automatically enters diff-aware mode when run on a feature branch to verify recent changes. The skill uses the Aside browser for interactive testing, captures screenshots as evidence, and produces structured reports with before/after health scores and a ship-readiness summary. It includes robust handling for AskUserQuestion decisions, including auto-decision rules for spawned sessions and prose fallbacks for interactive sessions, ensuring reliable operation across different execution contexts.
qa · high confidence
Introduce /setup-gbrain skill for coding-agent onboarding
Adds the \setup-gbrain\ skill, providing a guided, step-by-step workflow for users to install the gbrain CLI, initialize a local PGLite or Supabase database, or connect to a remote MCP brain, and register the tool with the coding agent. The skill includes robust state detection, broken-engine remediation, and detailed documentation on memory ingest, transcript handling, and per-repo trust policies to ensure secure and correct configuration.
setup-gbrain · high confidence
Introduce Codex skill with multi-mode review, challenge, and consult capabilities
Adds the Codex skill, enabling users to invoke the OpenAI Codex CLI for an independent second opinion via three modes: review (diff analysis with pass/fail gate), challenge (adversarial testing), and consult (general questions with session continuity). The skill includes a robust preamble that handles environment setup, degraded-mode fallbacks, and strict AskUserQuestion protocols (including prose fallbacks for Conductor sessions and auto-decision rules for spawned sessions) to ensure reliable user interaction across different execution contexts.
codex · high confidence
Introduce GStack Browser skill for AI-controlled Chromium automation
Users can now launch an AI-controlled Chromium browser instance via the new 'open-gstack-browser' skill, triggered by commands like 'open gstack browser' or 'launch chromium'. This skill provisions a visible, headed browser window with a built-in sidebar extension for real-time activity feeds and chat, featuring anti-bot stealth capabilities. The workflow handles setup, pre-flight cleanup of stale processes, connection verification on port 34567, and guides the user through pinning the extension, allowing for direct observation and control of browser actions by the AI agent.
open-gstack-browser · high confidence
Introduce dedicated design binary with persistent board daemon and vision-based quality gates
Adds a new stateless design CLI (\design/src/cli.ts\) for AI-powered UI mockup generation, featuring a persistent background daemon (\design/src/daemon.ts\) that hosts comparison boards for user review and feedback. The toolchain includes vision-based quality checks (\design/src/check.ts\) using GPT-4o to verify text readability and layout completeness, visual diffing (\design/src/diff.ts\) to compare mockups against live implementations, and a design-to-code prompt generator (\design/src/design-to-code.ts\). Authentication is handled via a new \design/src/auth.ts\ module that resolves API keys from \\~/.gstack/openai.json\ or environment variables, with specific safeguards to warn users if their key matches a local \.env\ file to prevent silent billing of other projects. The daemon client (\design/src/daemon-client.ts\) manages the lifecycle of the persistent board server, including identity verification and version mismatch handling to prevent accidental process termination or data loss. Also in this area: New browse daemon route table and security hardening.
design/src · high confidence
Introduce gstack browse Chrome extension with terminal, inspector, and ref overlays
This change adds the initial files for the 'gstack browse' Chrome extension (Manifest V3), providing a side panel that hosts an interactive Claude Code terminal (via xterm.js) and debug tools including a CSS inspector and element ref overlays. The extension connects to a local browse server on port 34567, handling authentication via a pinned extension identity and managing connection state through health polling. It includes a background service worker for command proxying and token management, content scripts for rendering ref badges and status pills on web pages, and a popup for port configuration.
extension · high confidence
Introduce make-pdf CLI for markdown-to-PDF conversion
This change adds the core implementation for the make-pdf tool, enabling users to convert Markdown documents into publication-quality PDFs. The new codebase includes a CLI entry point with argument parsing, a command registry, and an orchestrator that manages the rendering pipeline. Key capabilities include rendering Mermaid and Excalidraw diagrams, handling image inlining and size policies (including automatic landscape promotion for wide images), and supporting features like tables of contents, cover pages, watermarks, and page numbering. The tool uses the Aside browser (or a headless fallback) for rendering and includes a pdftotext wrapper for CI validation. Also in this area: Safer Git invocation and new PDF conversion type definitions.
make-pdf/src · high confidence
Introduce make-pdf skill for publication-quality PDF generation
Adds the make-pdf skill, enabling users to convert markdown files into high-quality PDFs with features like proper margins, page numbers, clickable tables of contents, and draft watermarks. The skill supports multiple output formats including HTML and DOCX, handles embedded diagrams (Mermaid, Excalidraw) and images, and intelligently manages font rendering for emoji and special characters across Linux, macOS, and Windows. It integrates with the gstack browser infrastructure, using the Aside browser when available or falling back to a headless browser, and includes robust error handling and CI-mode strictness for asset validation.
make-pdf · high confidence
Introduce offline diagram-render library with bundled Mermaid and Excalidraw
The \lib/diagram-render\ directory now contains a self-contained HTML page (\dist/diagram-render.html\) that bundles Mermaid, Excalidraw, and their converters for offline diagram rendering. This library provides a JavaScript API for rendering Mermaid diagrams to SVG, converting them to editable Excalidraw scenes, and rasterizing them to PNG, accessible via \lib/aside-render.ts\ and \bin/gstack-render.ts\. The bundle is committed to the repository to ensure zero-network rendering and includes pinned dependencies for Mermaid, Excalidraw, and React, along with a drift test to maintain build integrity.
lib/diagram-render · high confidence
Introduce plan-tune skill for question sensitivity and developer profile management
Adds the plan-tune skill, enabling users to tune how often gstack asks questions, set per-question preferences (never-ask, always-ask, or one-way), and view a dual-track developer profile that compares declared preferences against inferred behavior. The skill includes a consent gate, a 5-question setup wizard for initial profile declaration, and a 'dream cycle' feature to distill free-text feedback into actionable profile updates. It also defines strict formatting and fallback rules for AskUserQuestion prompts, including ELI10 explanations, completeness scores, and prose fallbacks for interactive sessions where tool calls fail.
plan-tune · high confidence
Introduce structured plan-eng-review sections and manifest
Added a new \plan-eng-review/sections\ directory containing a manifest (\manifest.json\) that registers review sections, alongside the \review-sections.md\ template (and generated output) defining the review preparation, write policies, confidence calibration, and decision procedures, and a \checklist.md\ template (and generated output) specifying the operative steps and required outputs for the \gstack-autoplan\ runner profile.
plan-eng-review/sections · high confidence
Introduces fail-closed freeze boundary enforcement hooks
Adds \check-freeze.sh\ and \freeze-state.sh\ to the \freeze/bin\ directory to enforce file-edit boundaries with strict security guarantees. The new \check-freeze.sh\ hook validates that tool inputs target files within the allowed directory, resolving symlinks to prevent boundary escapes and failing closed (denying access) if parsing fails, helpers are missing, or the state is ambiguous. The \freeze-state.sh\ script manages the boundary state file with atomic writes and mutex locking to prevent race conditions, ensuring that the freeze configuration is preserved or released safely. Also in this area: Unified state root for gstack data storage.
freeze/bin · high confidence
New 'claude-code' skill for external code reviews
A new skill definition (SKILL.md.tmpl) has been added to enable Claude Code to provide a second opinion on code changes when invoked from non-Claude Code hosts. This feature introduces three modes—review, challenge, and consult—allowing users to request diff reviews for bugs and security issues, adversarial testing for failure modes, or general repository consultation. The skill enforces strict execution boundaries, such as disabling tools in review/challenge modes, applying timeouts, and ensuring read-only access where appropriate, while integrating with the existing gstack runtime for invocation.
claude-code · high confidence
New 'health' skill for code quality dashboards
Added a new \health\ skill that acts as a code quality dashboard. It automatically detects or uses a configured 'Health Stack' of tools (type checkers, linters, test runners, dead code detectors, shell linters, and GBrain) to run checks, compute a weighted 0-10 composite score, and track trends. The skill handles complex decision flows via \AskUserQuestion\, including auto-decision logic for spawned/conductor sessions and prose fallbacks for interactive sessions, ensuring robust operation across different execution environments.
health · high confidence
New 'skillify' skill codifies scrape flows into permanent browser skills
A new 'skillify' skill (SKILL.md and SKILL.md.tmpl) allows users to convert the most recent successful /scrape flow into a permanent, deterministic browser skill on disk. This enables future /scrape calls with the same intent to run in \~200ms by executing the codified script instead of re-driving the page. The skill synthesizes a TypeScript script, a test file, and a fixture, runs the test in a temporary directory, and only commits the skill to disk after a passing test and explicit user approval, ensuring no broken skills are left on the system.
skillify · high confidence
New /benchmark skill for performance regression detection
A new \benchmark\ skill has been added to the gstack suite, enabling automated performance regression detection. It establishes baselines for page load times, Core Web Vitals (TTFB, FCP, LCP), and resource sizes, then compares current metrics against those baselines to identify regressions. The skill uses the Aside browser to collect real browser metrics and supports modes for capturing baselines (\--baseline\), quick checks (\--quick\), and diff-based benchmarking (\--diff\).
benchmark · high confidence
New /design-consultation skill for end-to-end design system creation
Introduces the /design-consultation skill, which guides users through creating a complete design system (aesthetic, typography, color, layout, spacing, motion) and generating a DESIGN.md file. The skill features a multi-phase workflow: gathering product context, optional competitive research via the Aside browser, and generating a full proposal with independent 'outside voices' from Codex and Claude. It includes robust anti-slop guardrails against common AI design patterns, font verification procedures, and a preview phase that generates either AI mockups or a self-contained HTML preview page for user approval.
design-consultation · high confidence
New /document-release skill for post-ship documentation updates
A new \/document-release\ skill has been added to handle post-ship documentation updates. It automatically analyzes code diffs to identify changes in features, behavior, and infrastructure, then cross-references these against existing documentation (README, ARCHITECTURE, CONTRIBUTING, CLAUDE.md) to detect drift and coverage gaps using a Diataxis-based audit lens. The skill polishes CHANGELOG voice, cleans up TODOs, and surfaces documentation debt in the PR body, while ensuring version bumps and risky narrative changes require explicit user confirmation via AskUserQuestion.
document-release · high confidence
New /freeze skill to restrict file edits to a specific directory
A new 'freeze' skill has been added that allows users to lock file editing to a specific directory for the current session. When invoked, the user specifies a target path, and the system installs PreToolUse hooks on the Edit and Write tools to check every operation against this boundary. Any attempt to edit or write a file outside the allowed directory is blocked (denied), while operations within the directory proceed normally. The restriction persists until explicitly removed via the /unfreeze command and is designed to prevent accidental changes to unrelated code during debugging or focused work. Also in this area: Unified state root for gstack data storage.
freeze · high confidence
New /guard skill combines destructive command warnings and directory-scoped edit restrictions
A new 'guard' skill has been added that activates both destructive command warnings (via the /careful hook) and directory-scoped edit restrictions (via the /freeze hook) in a single mode. Users can invoke this 'full safety mode' to protect against accidental deletions or modifications outside a specified directory boundary, with the system prompting for the target directory path upon activation.
guard · high confidence
New /setup-deploy skill for automated deployment configuration
A new 'setup-deploy' skill has been added to configure deployment settings for the /land-and-deploy workflow. This skill detects the user's deploy platform (Fly.io, Render, Vercel, Netlify, Heroku, GitHub Actions, or custom), identifies the production URL, health check endpoints, and deploy status commands. It then writes this configuration to CLAUDE.md, enabling automatic future deploys. The skill includes robust handling for plan mode, specific AskUserQuestion formatting for decision briefs, and fallback logic for degraded skill-start states.
setup-deploy · high confidence
New /unfreeze skill to clear edit restrictions
A new 'unfreeze' skill has been added to the system, allowing users to remove the directory edit restrictions previously set by the /freeze command without ending their session. This skill triggers on commands like 'unfreeze', 'unlock edits', or 'remove edit restrictions', and executes a script to clear the freeze state while logging the action to analytics. It ensures that existing /freeze hooks remain registered but effectively allow all edits until a new freeze is applied.
unfreeze · high confidence
New Aside browser driver and unified untrusted-content warning
The resolver layer now introduces a primary browser driver for Aside (scripts/resolvers/aside.ts) and a fallback headless browser driver (scripts/resolvers/browse.ts), along with composition helpers (scripts/resolvers/composition.ts) and confidence calibration (scripts/resolvers/confidence.ts). Aside is detected and driven via \aside repl\/\aside exec\ with strict rules (own tabs, consent for mutations, credential boundaries, one-flow-per-script), while the fallback uses gstack's own Chromium (\$B\) with a one-time build step. Both drivers share a single, centralized untrusted-content warning to prevent prompt injection drift. The composition resolver adds hooks for autoplan publication and CEO mode handoff, and an \INVOKE\_SKILL\ mechanism for composable skills. Confidence calibration enforces a pre-emit verification gate requiring quoted code lines for findings, suppressing low-confidence results from the main report.
scripts/resolvers · high confidence
New CEO review checklist and section manifest for unattended runs
The \plan-ceo-review/sections\ area now includes a \checklist.md\ (generated from \checklist.md.tmpl\) and a \manifest.json\ that registers the review sections. The checklist defines the specific steps and required outputs (such as reality rows, error/failure registries, and diagrams) for the \bin/gstack-autoplan\ runner in unattended or light modes, while the manifest provides a passive registry of section IDs, files, and scopes (e.g., \always\, \ui\) to guide the runner's section selection. The existing \review-sections.md\ (and its template) continues to provide the detailed 11-section deep review content for interactive runs.
plan-ceo-review/sections · high confidence
New DX review skill sections: checklist and 8-pass review guide
The \plan-devex-review/sections\ area now includes the core content files for the Developer Experience review skill. A new \manifest.json\ registers the skill sections, including a \runner\_only\ checklist for unattended or light \gstack-autoplan\ runs and the full interactive review guide. The \checklist.md\ (generated from \checklist.md.tmpl\) defines the DX phase steps and maps required outputs—such as the Developer Persona Card, Empathy Narrative, and DX Scorecard—to their producing sections. The \review-sections.md\ (generated from \review-sections.md.tmpl\) provides the operative 8-pass review guide (Getting Started, API/CLI/SDK Design, Error Messages, Documentation, Upgrade Path, Environment, Community, Measurement) with anti-skip rules, decision gates, and prior learning integration, ensuring consistent, thorough DX evaluation across all plan types.
plan-devex-review/sections · high confidence
New Hacker News front-page scraping skill
A new browser skill named \hackernews-frontpage\ has been added to the \browser-skills\ directory. When executed, it navigates to the Hacker News front page, parses the top 30 stories, and outputs a JSON document containing each story's rank, title, URL, point count, and comment count. The implementation includes a dedicated parser script, a bundled HTML fixture for testing, and a test suite that validates the parsing logic against the fixture without requiring a live network connection.
browser-skills · high confidence
New QA skill sections for browser setup, verification, and exploratory testing
The QA skill now includes a structured set of documentation sections in \qa/sections/\ to guide browser-based and functional quality assurance. \browser-setup.md\ defines how to use the Aside browser or a headless fallback, including strict rules for tab isolation, consent for mutations, and credential handling. \browser-verify.md\ provides a protocol for re-running interactions to verify browser defect repairs. \exploratory.md\ outlines the shared exploratory QA loop, including chartering, deadline-bound probing, and evidence recording. \qa-patterns.md\ details the browser QA methodology, covering diff-aware, full, quick, and regression modes, along with a Health Score Rubric. \system-functional.md\ specifies testing for APIs, CLIs, and jobs using native repository tools. \test-bootstrap.md\ guides the automated detection and installation of test frameworks for browser targets. \scope.md\ establishes rules for selecting QA surfaces and ensuring isolation. A \manifest.json\ registers these sections for automatic loading based on the QA context.
qa/sections · high confidence
New browse skill documentation for command reference and snapshot flags
Added a generated command reference document (\command-list.md\) and its template (\command-list.md.tmpl\) to the browse skill sections, covering snapshot flags, navigation, reading, extraction, interaction, inspection, visual, and meta commands. A manifest entry registers this section to be loaded when users invoke browse commands or snapshot flags beyond the basic fallback translation table, ensuring the full command syntax and argument shapes are available to the agent.
browse/sections · high confidence
New browser-skill infrastructure with secure, scoped execution and auditability
This change introduces the core backend for the browser-skill system in browse/src, enabling agents and users to install, run, and manage Playwright-based automation scripts. It adds a scoped authentication model via browse-client.ts, which resolves daemon tokens from environment variables or a state file to ensure skills only access the browser with limited, per-spawn permissions. Security is enforced through browser-skill-commands.ts, which spawns scripts with scrubbed environments and locked working directories for untrusted skills, and browser-skill-write.ts, which uses atomic staging and commit logic to prevent partial or malicious skill installations. The system also includes robust operational features: activity.ts provides a real-time, privacy-filtered (redacting passwords/tokens) streaming feed of browser commands for the Chrome extension, while audit.ts maintains a persistent, append-only JSONL log of all commands for forensic review. Finally, cdp-allowlist.ts and cdp-bridge.ts implement a strict default-deny policy for Chrome DevTools Protocol access, ensuring only audited, safe methods are exposed to skills.
browse/src · high confidence
New cross-model benchmark skill for comparing AI providers
Added a new \benchmark-models\ skill that allows users to run side-by-side comparisons of AI models (Claude, GPT, Gemini) on specific prompts or existing skills. The skill guides users through selecting a prompt, verifying provider authentication via a dry-run, optionally enabling a quality judge, and executing the benchmark to compare latency, cost, and output quality. It includes logic to handle plan mode, telemetry logging, and saving results for future trend analysis.
benchmark-models · high confidence
New design-consultation section for proposal and preview phases
Added a new 'proposal-and-preview' section to the design-consultation skill, including a manifest entry and detailed procedural documentation. This section guides the system through Phase 3 (complete design proposal) and Phase 5 (design system preview), enforcing specific aesthetic directions, font verification procedures, and anti-slop constraints. It also introduces logic for optional 'outside voices' (independent design reviews from Codex) and defines fallback paths for generating AI mockups or HTML previews when design tools are unavailable.
design-consultation/sections · high confidence
New design-html sections: UX doctrine, Pretext API patterns, and detector install offer
The design-html skill now includes three new documentation sections to guide its behavior. First, a UX principles doctrine defines usability laws (e.g., 'don't make me think', 'omit, then omit again') and behavioral observations (scanning, satisficing) that govern all design decisions. Second, a Pretext wiring patterns reference provides four specific JavaScript usage patterns for the \@chenglou/pretext\ library—covering basic height computation, tight-fit containers, text around obstacles, and line-by-line rendering—along with a complete API cheatsheet. Third, a one-time install offer section handles the user consent flow for downloading the 'impeccable' design detector engine, specifying the exact question to ask, the installation command, and configuration options for enabling or disabling the detector. Also in this area: Persisted design archetypes and craft exceptions flow from design-shotgun to design-html.
design-html/sections · high confidence
New design-html skill for Pretext-native HTML generation
The design-html skill is now available to generate production-quality HTML/CSS using the Pretext engine for dynamic text layout. It accepts input from approved mockups, CEO plans, or direct user descriptions, and includes a one-time offer to install the 'impeccable' design detector engine for automated CSS validation. The skill provides specific JavaScript patterns for height computation, tight-fit containers, and editorial layouts, ensuring text reflows and heights adjust correctly to content.
design-html · high confidence
New design-review skill for visual QA and automated fixes
A new \design-review\ skill (v2.0.0) has been added to the gstack suite. It performs visual design audits—checking for inconsistencies, spacing issues, hierarchy problems, and AI-generated patterns—and iteratively fixes them in source code with atomic commits and before/after verification. The skill supports plan-mode reviews, integrates with the Aside browser for live site inspection, and uses a structured AskUserQuestion format for decision briefs, including prose fallbacks for interactive sessions and auto-decision rules for spawned or conductor sessions.
design-review · high confidence
New design-shotgun skill for visual design exploration and iteration
Introduces the \design-shotgun\ skill, enabling users to generate multiple AI design variants, view them side-by-side in a comparison board, and collect structured feedback to iterate on visual directions. The skill includes a preamble for session detection and degraded-mode handling, a robust \AskUserQuestion\ protocol with prose fallbacks for decision briefs, and a 'taste memory' system that learns from approved/rejected variants to bias future generations. It also incorporates a UX principles doctrine to guide design decisions based on observed user behavior.
design-shotgun · high confidence
New devex-review skill for live developer experience audits
A new \devex-review\ skill has been added to the gstack suite, enabling live, automated testing of the developer experience. This skill navigates documentation and getting-started flows in the Aside browser, times to-first-helpful-work (TTHW), captures screenshots of error messages, and evaluates CLI help text to produce a scored DX scorecard with evidence. It includes a 'boomerang' feature that compares current live results against prior \/plan-devex-review\ predictions, and provides detailed instructions for handling AskUserQuestion interactions, including prose fallbacks for constrained environments like Conductor sessions.
devex-review · high confidence
New diagram skill for generating editable diagram triplets
A new 'diagram' skill has been added to the system, enabling users to generate a complete diagram triplet (Mermaid source, editable Excalidraw file, and rendered SVG/PNG) from an English description or Mermaid code. The skill operates fully offline using a self-contained render bundle and supports both flowcharts/sequence diagrams (which convert to editable Excalidraw scenes) and other diagram types. It includes robust error handling for Mermaid parse errors and fallbacks for missing browsers or bundles, ensuring users always receive valid, editable artifacts.
diagram · high confidence
New document-generate skill for creating structured documentation
Added the /document-generate skill, which uses the Diataxis framework to produce tutorials, how-tos, references, and explanations for features, modules, or entire projects. The skill supports standalone invocation or integration with /document-release to fill coverage gaps, and includes robust handling for plan mode, spawned sessions, and AskUserQuestion failures via prose fallbacks.
document-generate · high confidence
New documentation audit sections for release workflows
Added new documentation sections (\audit-scope.md\, \release-body.md\) and their templates to define the behavior of the document-release skill. These files introduce a 'ship-owned' mode that allows subagent dispatches to perform scoped documentation audits—such as per-file checks, CHANGELOG voice polishing, and cross-document consistency—without stranding the run or mutating Git metadata. A new \manifest.json\ registers these sections, enabling the system to discover and execute these specific audit steps during the release process.
document-release/sections · high confidence
New gstack-upgrade skill for safe, automated version updates
A new skill named 'gstack-upgrade' (v1.1.0) has been added to handle upgrading the gstack installation. It supports both git-based and vendored install types, automatically detecting the current setup to apply the correct upgrade logic. The skill features an inline upgrade flow that respects user preferences, including an auto-upgrade mode, snooze functionality with escalating backoff (24h, 48h, 1 week), and the ability to disable update checks entirely. For git installs, it prioritizes fast-forward merges and safely handles local changes by stashing or requiring explicit user confirmation before destructive resets. For vendored installs, it uses a backup-and-replace strategy with automatic restoration on failure, ensuring users are not left with a broken installation. Also in this area: Upgrade process now supports stable update channels and fixes Windows shell configuration; Fixes to autoplan, CI gates, nested-group state, and gbrain context loading; Unified state root for gstack data storage.
gstack-upgrade · high confidence
New iOS Debug Bridge templates for automated QA
Added a suite of Swift and Objective-C templates to the ios-qa toolchain that generate a local debug server and UI overlay for iOS app testing. The generated code includes a loopback StateServer for reading and writing app state, a touch-synthesis engine for reliable SwiftUI and UIKit interaction, and a visual DebugOverlay to indicate active testing sessions. These templates are strictly gated to DEBUG builds to prevent private API exposure in release binaries.
ios-qa/templates · high confidence
New iOS QA skill for live-device testing
Adds a new \ios-qa\ skill that enables automated quality assurance on real iPhone and iPad devices. The skill connects to a device via USB CoreDevice, reads Swift source code to understand the app's UI state, and runs a vision-driven agent loop (screenshot, analyze, decide, act, verify) to find and fix bugs. It supports remote testing via Tailscale and includes detailed instructions for bootstrapping a debug bridge, managing session tokens, and handling user questions through the AskUserQuestion tool or prose fallbacks.
ios-qa · high confidence
New iOS design review and autonomous bug-fix skills
Two new gstack skills are introduced to close the loop on iOS visual quality and defects. The ios-design-review skill connects to a real iPhone via the gstack StateServer to screenshot every screen and evaluate it against Apple HIG, DESIGN.md, and best practices, scoring ten dimensions (typography, spacing, color, touch targets, loading/empty/error states, accessibility, animation, iOS idioms, information density, and AI-slop) on a 0–10 scale with actionable feedback. The ios-fix skill takes a bug reported by /ios-qa, captures a reproducing state snapshot as a regression fixture, locates the root cause in Swift source, applies a minimal fix, rebuilds and redeploys to the device, verifies the fix against the snapshot, and adds a regression test in test/fixtures/ios-fix. Both skills use the same preamble protocol (gstack-skill-start) for session/onboarding/telemetry, support plan mode with AskUserQuestion decision briefs (including prose fallbacks and spawned/Conductor handling), and are triggered by voice/text commands such as 'review the iOS design' and 'fix this iOS bug'.
ios-design-review · high confidence
New interactive developer experience (DX) plan review skill
Adds the \/plan-devex-review\ skill, an interactive tool for auditing developer-facing plans (APIs, CLIs, SDKs, libraries, platforms, and docs). The skill guides users through a structured review process—starting with developer persona identification and competitive benchmarking, then moving through eight specific review passes (Getting Started, API/CLI Design, Error Messages, Documentation, Migration Paths, Tooling, Community, and Measurement). It enforces rigorous decision-cadence rules using \AskUserQuestion\ with structured decision briefs (including ELI10 explanations, completeness scores, and pros/cons) to ensure explicit user approval for scope and design choices. The skill includes a reference \dx-hall-of-fame.md\ with industry gold standards and anti-patterns to ground the review in evidence rather than opinion.
plan-devex-review · high confidence
New interactive plan-design-review skill for pre-implementation design critique
A new /plan-design-review skill has been added, allowing users to review design plans before implementation. The skill acts as a senior product designer, rating design dimensions on a 0-10 scale, identifying missing decisions, and suggesting fixes. It features a strict scope gate to confirm the review target (branch diff, plan doc, or specific path), integrates with the gstack designer to generate visual mockups for UI plans, and enforces a structured decision-brief format for user interactions. The skill is triggered by phrases like 'design plan review' or 'review ux plan' and is designed to work in plan mode, ensuring design quality is addressed early in the development cycle.
plan-design-review · high confidence
New ios-clean skill to remove DebugBridge instrumentation
Added a new \ios-clean\ skill (SKILL.md and template) that provides a guided workflow to remove the DebugBridge SPM package, \#if DEBUG wiring, StateServer/DebugOverlay artifacts, and generated StateAccessor files from an iOS app. The skill handles inventory, removal, and verification (including a Release build check) while deferring to the structural Package.swift guard for safety, and integrates with the gstack preamble for session context and AskUserQuestion decision flows.
ios-clean · high confidence
New ios-sync skill for regenerating the iOS debug bridge
Added the ios-sync skill (SKILL.md and SKILL.md.tmpl) which allows users to regenerate the iOS debug bridge accessors and instrumentation against the latest upstream gstack templates. This skill handles detecting installed versions, running the deterministic regenerator to update StateServer.swift, DebugOverlay.swift, Package.swift, and typed @Observable state accessors, and verifying the build, ensuring the iOS debug bridge stays in sync with gstack upgrades or new ViewModels.
ios-sync · high confidence
New land-and-deploy skill for automated PR merging and production verification
Introduces the land-and-deploy skill, which automates the workflow of merging a Pull Request, waiting for CI and deployment, and verifying production health via canary checks. The skill triggers on commands like 'merge', 'land', or 'deploy', taking over after the /ship command creates the PR. It includes a robust preamble that initializes the gstack environment, handles degraded modes if the start script is unavailable, and manages user interactions through a structured AskUserQuestion format with prose fallbacks for interactive sessions. The workflow enforces pre-flight checks (authentication, PR state, branch matching), pre-merge validation (CI status, conflicts), and a first-run dry-run validation to confirm deployment configuration before proceeding. It specifically notes that GitLab support is not yet implemented, requiring manual merging in that case.
land-and-deploy · high confidence
New land-and-deploy skill sections for first-run validation, readiness gating, and merge/deploy
Added the core documentation sections for the land-and-deploy skill: a first-run validation step (1.5) that detects deploy infrastructure and validates configuration before any merge; a pre-merge readiness gate (3.5) that checks review staleness, test results, and PR body accuracy; and a merge-and-deploy section (4) that handles PR merging, merge-queue detection, and CI auto-deploy monitoring. These files define the decision-tree logic and shell commands the skill uses to safely land and deploy code.
land-and-deploy/sections · high confidence
New landing-report skill for version queue visibility
Added a new \landing-report\ skill that provides a read-only dashboard of the version queue. It shows which version slots are claimed by open PRs, identifies sibling Conductor workspaces with work likely to ship soon, and indicates which version slot \/ship\ would claim next. The skill uses the \gstack-next-version\ utility to query micro, patch, minor, and major bump levels, rendering a table of claimed versions, active siblings, and potential collisions, with offline fallbacks when the queue is unavailable.
landing-report · high confidence
New pair-agent skill for remote browser sharing
Added a new 'pair-agent' skill that allows a local AI agent to share its browser session with a remote AI agent (such as OpenClaw, Hermes, Codex, or Cursor). The skill generates a one-time setup key and provides instructions for the remote agent to connect via an ngrok tunnel (for remote machines) or direct config write (for local machines), enabling the remote agent to browse the web using the local browser's tabs and cookies.
pair-agent · high confidence
New plan-design-review sections: checklist and review-pass definitions
The \plan-design-review/sections\ area now includes a \manifest.json\ that registers two new review sections: \review-sections.md\ (defining the seven design passes and required outputs for interactive runs) and \checklist.md\ (providing a runner-only profile for unattended or light \gstack-autoplan\ runs). The \checklist\ section maps specific outputs—such as NOT in scope, What already exists, TODOS.md updates, Completion Summary, Unresolved Decisions, Approved Mockups, Review log row, and Next-skill handoff—to their producing sections, ensuring the unattended runner knows exactly which artifacts to generate and where they belong.
plan-design-review/sections · high confidence
New report-only QA skill for bug detection without code changes
A new 'qa-only' skill has been added to the system, allowing users to request a structured QA report (including health scores, screenshots, and repro steps) without triggering any code fixes. This skill is triggered by phrases like 'just report bugs' or 'qa report only' and is designed to run in diff-aware mode by default when on a feature branch, producing reports in \.gstack/qa-reports/\. It operates as a read-only counterpart to the full \/qa\ skill, ensuring that testing activities do not inadvertently modify the codebase.
qa-only · high confidence
New runtime qualification infrastructure for Node, Bun, Python, and Rails
The \lib/cso/images\ directory now contains the foundational recipes and configuration for a new runtime qualification system. This includes pinned \build-inputs.json\ defining reviewed base images for Node (24.4.0), Bun (1.3.10), Python (3.13.4 with uv 0.8.0), and Rails (Ruby 3.4.4), alongside specific Dockerfiles and helper scripts (entrypoint, verifier, preparation) to ensure reproducible, secure builds. A \qualification.json\ manifest enforces strict security and functional checks—such as offline execution, non-root isolation, and held-out repair testing—before a runtime profile is considered qualified, addressing the need for native runtime qualification across these platforms.
lib/cso/images · high confidence
New scrape skill for read-only web data extraction via Aside browser
A new \scrape\ skill has been added to the gstack suite, enabling the system to pull structured data from web pages using the user's existing Aside browser sessions. The skill is strictly read-only; it refuses any intent that implies mutating actions (such as form submissions or logins) and instead directs users to the \/qa\ flow for such tasks. It supports both structured extraction (via DOM selectors) and fuzzy queries (via Aside's built-in agent), returning results as a single JSON document on stdout. The skill includes robust safety measures, such as treating all page content as untrusted input, handling sign-in walls by prompting the user to log in manually, and ensuring browser tabs are closed automatically to leave the browser state unchanged.
scrape · high confidence
New skill to import browser cookies for authenticated headless browsing
Added a new \setup-browser-cookies\ skill that allows users to import cookies from their real Chromium browser into the headless browse session. This enables QA testing of authenticated pages by providing an interactive picker to select cookie domains and accounts. The skill handles browser setup verification, safe cookie extraction with explicit user consent for options like clearing storage, and reports import status honestly without exposing sensitive data.
setup-browser-cookies · high confidence
New sync-gbrain skill to keep the agent's code search index current
A new \sync-gbrain\ skill has been added to allow users to manually refresh the gbrain code index and update agent search guidance. Triggered by commands like \/sync-gbrain\, \sync gbrain\, or \reindex repo\, this skill wraps the \gstack-gbrain-sync\ orchestrator to probe state, register native code surfaces, and reindex the repository. It supports various modes including incremental syncs, full reindexes, and call-graph generation, while handling local engine pre-flight checks and remote brain configurations to ensure the coding agent has up-to-date context for code search.
sync-gbrain · high confidence
Office Hours skill adds structured diagnostic and design doc sections
The office-hours skill now includes a manifest and dedicated sections for Phase 2A (startup diagnostic), Phase 2B (builder brainstorm), and Phase 5 (design doc and handoff). The startup diagnostic enforces rigorous, evidence-based questioning on demand and user specificity, while the builder section focuses on creative ideation. The design doc section introduces a dual-write mechanism that saves a design document to both a private state directory and the project repository, including automatic lineage tracking and redaction checks before committing to the repo.
office-hours/sections · high confidence
Office-hours skill introduces dual-mode design workflow with security-scanned repo handoff
The office-hours skill now operates in two distinct modes—Startup (for founders) and Builder (for side projects/hackathons)—using a structured diagnostic process to generate a design document. A key behavioral change is the introduction of a dual-write mechanism for the final design doc: it saves a private copy in \~/.gstack for cross-session memory and, when run inside a git repository, writes a team-shareable copy to docs/designs/. This public-facing copy is strictly gated by a pre-write security scan (gstack-redact) to prevent accidental leakage of secrets, with a non-blocking fallback if the scan fails or the repo is read-only. The skill also implements a tiered relationship closing that adapts its tone based on the user's session history.
office-hours · high confidence
Post-deploy canary monitoring skill
A new 'canary' skill has been added to the gstack suite to provide post-deploy visual monitoring. It watches the live application for console errors, performance regressions, and page failures by taking periodic screenshots and comparing them against pre-deploy baselines. The skill supports user-invocable commands like \/canary \<url\>\ with options for custom durations, baseline capture, and specific page monitoring. It automatically discovers pages, captures baselines, and alerts on anomalies such as new console errors, load time regressions, or broken links, ensuring reliability between deployment and verification.
canary · high confidence
Review skill restructured with Fix-First workflow and expanded checklist
The review skill has been significantly expanded and restructured. The main SKILL.md now includes a comprehensive AskUserQuestion format with ELI10 explanations, completeness scoring, and support for split questions. The workflow now implements a 'Fix-First Review' approach where mechanical fixes are applied automatically and ambiguous issues are batched into single user questions. The checklist has been updated with new critical categories including Race Conditions & Concurrency, Shell Injection (Python-specific), and Enum & Value Completeness. Additional features include Greptile comment triage integration, workspace-aware queue status checking, and slop scanning for AI code quality issues. The skill now supports parallel specialist reviewers for categories like Test Gaps and Dead Code.
review · high confidence
Structured setup-gbrain sections for engine remediation, brain initialization, and transcript consent
The setup-gbrain skill now uses a modular section system (managed by sections/manifest.json) to handle distinct setup phases. This introduces a Step 1.5 remediation flow that detects broken local engines and offers users a rollback-safe switch to local PGLite or a retry. The brain initialization (Step 4) is now path-specific, supporting Supabase (existing URL, auto-provision, or manual), local PGLite, and remote MCP connections, with explicit PAT scope disclosures and TLS requirements for remote setups. Additionally, a new transcript consent gate (Step 7.5) allows users to opt-in to ingesting coding-agent session transcripts into their brain, with options to scope by time window and repository.
setup-gbrain/sections · high confidence
iOS QA daemon introduces secure remote access with capability-based authentication and session persistence
The iOS QA daemon now supports secure remote access over Tailscale, gated by a capability-tier allowlist and short-lived session tokens. Remote agents must be explicitly granted access via the \gstack-ios-qa-mint\ CLI (owner-granted) or by being pre-authorized in the allowlist (self-service mint), with capabilities ranging from 'observe' to 'restore'. The daemon persists the rotated boot token to disk, allowing a restarted daemon to resume the device session without relaunching the app and losing QA state. It also includes robust tunnel management, including keepalive mechanisms to prevent CoreDevice tunnel drops, and comprehensive audit logging for authentication attempts.
ios-qa/daemon · high confidence
Behavioural changes
1 commit (0 fixes) modifying patches
A change to existing behaviour in patches — 1 commit, 1 file: “Playwright 1.65.0: bundled Chromium 158 drops the distrusted eMudhra root (\#3338) (\#4087)” (refs \#3338, \#4087).
patches · medium confidence · unverified
119 commits (56 fixes) modifying ship/sections
A change to existing behaviour in ship/sections — 119 commits (56 fixs), 34 files: “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “v1.91.2.0 fix: consolidate gstack reliability wave (\#2959)”; “v1.81.0.0 feat: Aside is the browser gstack drives first; every browsing skill, the PDF/diagram renderer, and web research; the bundled browser stays the automatic fallback (\#2810)”; and 116 more (refs \#2571, \#2959, \#2810, \#2925, \#3339, …).
ship/sections · medium confidence · unverified
13 commits (3 fixes) modifying captain
A change to existing behaviour in captain — 13 commits (3 fixs), 2 files: “feat(captain): postmortem template and step, first on \#3787; Unverified line in READY and lane reports (GST-805) (\#3807)”; “v1.91.276.0 feat(preamble): tier 2+ skills load the shared preamble once per session (GSTA-96) (\#3434)”; “v1.92.12.0 feat: wave PRs land 3 to 6 queued PRs as one train slot (gstack-ship-queue wave) (GST-574) (\#3772)”; and 10 more (refs \#3787, \#3807, \#3434, \#3772, \#3473, …).
captain · medium confidence · unverified
135 commits (45 fixes) modifying autoplan
A change to existing behaviour in autoplan — 135 commits (45 fixs), 5 files: “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “v1.91.240.0 feat(proactive): default to the four core skills (core \| all \| off), with docs, prompt and tests (GSTA-99) (\#3432)”; “v1.29.0.0 feat: worktree-aware gbrain code sources via path-hash IDs and CWD pin (\#1382)”; and 132 more (refs \#2571, \#3432, \#1382, \#2925, \#2722, …).
autoplan · medium confidence · unverified
14 commits (2 fixes) modifying eval-plan
A change to existing behaviour in eval-plan — 14 commits (2 fixs), 3 files: “v1.91.240.0 feat(proactive): default to the four core skills (core \| all \| off), with docs, prompt and tests (GSTA-99) (\#3432)”; “v1.91.373.0 feat(eval-plan): require a no-gstack arm in every head-to-head preregistration (\#3690)”; “v1.91.276.0 feat(preamble): tier 2+ skills load the shared preamble once per session (GSTA-96) (\#3434)”; and 11 more (refs \#3432, \#3690, \#3434, \#3708, \#3442, …).
eval-plan · medium confidence · unverified
17 commits (7 fixes) modifying test-audit
A change to existing behaviour in test-audit — 17 commits (7 fixs), 3 files: “v1.91.240.0 feat(proactive): default to the four core skills (core \| all \| off), with docs, prompt and tests (GSTA-99) (\#3432)”; “v1.91.32.0 fix: Codex works from zsh, free text never enters shell commands, honest failures (\#3057)”; “feat: cut standalone /plan-eng-review; /autoplan keeps its eng phase (GST-1698) (\#4108)”; and 14 more (refs \#3432, \#3057, \#4108, \#3434, \#3115, …).
test-audit · medium confidence · unverified
2 commits (0 fixes) modifying crew/scripts
A change to existing behaviour in crew/scripts — 2 commits, 14 files: “Stop the stable cut counting its own check runs against the commit it cuts (GST-2463) (\#4297)”; “Move the tools only Capy crews use into a crew/ folder (GST-1833) (\#4145)” (refs \#4297, \#4145).
crew/scripts · medium confidence · unverified
281 commits (104 fixes) modifying scripts
A change to existing behaviour in scripts — 281 commits (104 fixs), 97 files: “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “feat: /land-and-deploy, /canary, /benchmark + perf review (v0.7.0) (\#183)”; “feat: review chaining + commit hash staleness tracking (v0.8.3) (\#206)”; and 278 more (refs \#2571, \#183, \#206, \#2959, \#3432, …).
scripts · medium confidence · unverified
3 commits (0 fixes) modifying design-motion
A change to existing behaviour in design-motion — 3 commits, 10 files: “v1.91.276.0 feat(preamble): tier 2+ skills load the shared preamble once per session (GSTA-96) (\#3434)”; “v1.93.6.0 feat: /plan-tune learns from ad-hoc question ids that carry a signal hint (\#3442) (\#3868)”; “v1.91.264.0 feat(design-motion): new /design-motion skill for showcase loops, adapted from Elaya's design-to-motion (GSTA-92 phase 2) (\#3430)” (refs \#3434, \#3442, \#3868, \#3430).
design-motion · medium confidence · unverified
3 commits (0 fixes) modifying verify
A change to existing behaviour in verify — 3 commits, 2 files: “v1.94.22.0 feat(skills): when-to-use descriptions, wider triggers and contract blocks for /investigate, /review, /qa, /verify (GST-1088 lane B) (\#3909)”; “v1.91.359.0 feat(verify): /verify skill and gstack-checks bin (GST-116 PR 1) (\#3672)”; “feat: retire /investigate after a no-gain eval (GST-1717) (\#4110)” (refs \#3909, \#3672, \#4110).
verify · medium confidence · unverified
4 commits (0 fixes) modifying crew/lib
A change to existing behaviour in crew/lib — 4 commits, 29 files: “Read the merge queue's notes in one git process so land and ready stay fast (GST-2198) (\#4242)”; “Warn when a branch-only run hides a red PR check, and make ready refuse when its note or base is missing (GST-2205) (\#4270)”; “Move the tools only Capy crews use into a crew/ folder (GST-1833) (\#4145)”; and 1 more (refs \#4242, \#4270, \#4145, \#4254).
crew/lib · medium confidence · unverified
4 commits (0 fixes) modifying write-skill
A change to existing behaviour in write-skill — 4 commits, 4 files: “v1.91.318.0 test(skill-evals): office-hours eval tasks for the shared skill eval; /write-skill diet locks tasks before a paid run (GSTA-142) (\#3550)”; “v1.93.6.0 feat: /plan-tune learns from ad-hoc question ids that carry a signal hint (\#3442) (\#3868)”; “v1.91.287.0 feat(write-skill): write, change and diet skills; /learn at half its template size (\#3524)”; and 1 more (refs \#3550, \#3442, \#3868, \#3524, \#3837).
write-skill · medium confidence · unverified
40 commits (15 fixes) modifying ios-fix
A change to existing behaviour in ios-fix — 40 commits (15 fixs), 3 files: “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “v1.91.240.0 feat(proactive): default to the four core skills (core \| all \| off), with docs, prompt and tests (GSTA-99) (\#3432)”; “v1.89.0.0 feat: add shared-code extraction audit (\#2925)”; and 37 more (refs \#2571, \#3432, \#2925, \#2722, \#2078, …).
ios-fix · medium confidence · unverified
40 commits (19 fixes) modifying review/sections
A change to existing behaviour in review/sections — 40 commits (19 fixs), 10 files: “v1.91.2.0 fix: consolidate gstack reliability wave (\#2959)”; “v1.89.0.0 feat: add shared-code extraction audit (\#2925)”; “v1.75.0.0 feat: ponytail import wave — simplification review lens, arm benchmark, reuse ladder, instruction-tier digest (\#2722)”; and 37 more (refs \#2959, \#2925, \#2722, \#497, \#2772, …).
review/sections · medium confidence · unverified
446 commits (218 fixes) modifying lib
A change to existing behaviour in lib — 446 commits (218 fixs), 256 files: “v1.91.186.0 fix(redact): pii.cc skips compact YYYYMMDDHHMMSS stamps so autoplan approval imports stop failing 1 in 10 (GSTA-66) (\#3366)”; “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “Make the ship queue fetch the latest main before it checks a pull request (GST-2232) (\#4256)”; and 443 more (refs \#3366, \#2571, \#4256, \#2959, \#3931, …).
lib · medium confidence · unverified
483 commits (231 fixes) modifying bin
A change to existing behaviour in bin — 483 commits (231 fixs), 201 files: “v1.92.48.0 fix(windows): review-scope summary runs gstack-review-read through Git Bash; jsonl-merge writes UTF-8 with LF (\#3541, \#3659) (\#3714)”; “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “Make the ship queue fetch the latest main before it checks a pull request (GST-2232) (\#4256)”; and 480 more (refs \#3541, \#3659, \#3714, \#2571, \#4256, …).
bin · medium confidence · unverified
5 commits (1 fix) modifying lib/design-lint
A change to existing behaviour in lib/design-lint — 5 commits (1 fix), 31 files: “Add /design-review --build, an opt-in loop that checks and fixes new UI while you build it (\#4142)”; “Let the design lint check one section of a running app for crowded rows, cramped help text, loud notes and small tap targets (\#4138)”; “v1.91.260.0 feat(design-review,plan-design-review): craft audit and Persuade plan checks; craft tells as checker-only catalog entries (GSTA-92 phase 2) (\#3433)”; and 2 more (refs \#4142, \#4138, \#3433, \#3507, \#3459).
lib/design-lint · medium confidence · unverified
54 commits (24 fixes) modifying lib/pregate
A change to existing behaviour in lib/pregate — 54 commits (24 fixs), 26 files: “ship accept: a lanes-only head policy gates the final tree; owner-granted test-lane exception (GST-1595) (\#4061)”; “v1.91.364.0 fix: pregate --policy-from head only adds lanes, and timeout\_s reaches preflight and finish (GST-419, GST-420) (\#3676)”; “v1.91.107.0 fix: the GBRA-40 field list (one Bun for install and check, \rec\ gate approval, pregate never silent) (\#3190)”; and 51 more (refs \#4061, \#3676, \#3190, \#3747, \#3960, …).
lib/pregate · medium confidence · unverified
6 commits (0 fixes) modifying modernize
A change to existing behaviour in modernize — 6 commits, 3 files: “v1.91.284.0 feat(absorb-tstack): /modernize skill, per-skill Codex default prompts, gstack-post --attribute, the unattended grant and a CONTRIBUTING porting section (\#3497)”; “feat(hosts): /land-and-deploy and /modernize run only when typed on Claude Code and Cursor (C1) (\#3808)”; “docs: on Capy, comment and request reviews only through tools.pr.comment / tools.pr.edit, never gh (GST-1417 follow-up) (\#4027)”; and 3 more (refs \#3497, \#3808, \#4027, \#3442, \#3868, …).
modernize · medium confidence · unverified
6 commits (1 fix) modifying design-review/sections
A change to existing behaviour in design-review/sections — 6 commits (1 fix), 10 files: “Add /design-review --build, an opt-in loop that checks and fixes new UI while you build it (\#4142)”; “v1.91.276.0 feat(preamble): tier 2+ skills load the shared preamble once per session (GSTA-96) (\#3434)”; “v1.91.260.0 feat(design-review,plan-design-review): craft audit and Persuade plan checks; craft tells as checker-only catalog entries (GSTA-92 phase 2) (\#3433)”; and 3 more (refs \#4142, \#3434, \#3433, \#4140, \#3525, …).
design-review/sections · medium confidence · unverified
6 commits (2 fixes) modifying capy/skill/gstack/scripts
A change to existing behaviour in capy/skill/gstack/scripts — 6 commits (2 fixs), 1 file: “v1.91.107.0 fix: the GBRA-40 field list (one Bun for install and check, \rec\ gate approval, pregate never silent) (\#3190)”; “v1.91.258.0 feat(capy-install): project floor, --for none, WARN for behind and Node-only pins, in-place review-voice upgrade, /cso env hygiene (GSTA-114 lane F, PR 1) (\#3475)”; “setup: new installs default to stable; Capy threads run latest main daily (GST-1642) (\#4091)”; and 3 more (refs \#3190, \#3475, \#4091, \#3164, \#3234, …).
capy/skill/gstack/scripts · medium confidence · unverified
6 commits (2 fixes) modifying runaway-stopper
A change to existing behaviour in runaway-stopper — 6 commits (2 fixs), 4 files: “feat: /runaway-stopper — detect, stop and prevent runaway threads and subagents (GST-1279) (\#3975)”; “Keep crew threads on their PR to the end and cap finished PRs at 10 before new fix threads (GST-2192) (\#4240)”; “Stop the runaway scan from telling Captains to rotate crew threads that hold a PR (GST-2226) (\#4250)”; and 3 more (refs \#3975, \#4240, \#4250, \#4244, \#3983, …).
runaway-stopper · medium confidence · unverified
7 commits (3 fixes) modifying capy
A change to existing behaviour in capy — 7 commits (3 fixs), 2 files: “v1.91.332.0 feat(capy): remember a user's model preference (agent\_model); every Capy session reads factory.md (\#3634)”; “v1.91.101.0 fix: close the field wave's residual gaps (\#3173)”; “Move the tools only Capy crews use into a crew/ folder (GST-1833) (\#4145)”; and 4 more (refs \#3634, \#3173, \#4145, \#3232, \#3475, …).
capy · medium confidence · unverified
7 commits (3 fixes) modifying lib/gate-diff
A change to existing behaviour in lib/gate-diff — 7 commits (3 fixs), 6 files: “v1.93.33.0 wave(review): \#3705 \#3744 \#3745 \#3764 \#3765 \#3795 (\#3852)”; “Stop the focused-test tag from keeping run data in review (GST-2006) (\#4195)”; “v1.92.76.0 wave(review): \#3809 \#3813 \#3826 \#3831 \#3838 (\#3848)”; and 4 more (refs \#3705, \#3744, \#3745, \#3764, \#3765, …).
lib/gate-diff · medium confidence · unverified
81 commits (34 fixes) modifying scripts/lib
A change to existing behaviour in scripts/lib — 81 commits (34 fixs), 15 files: “v1.92.48.0 fix(windows): review-scope summary runs gstack-review-read through Git Bash; jsonl-merge writes UTF-8 with LF (\#3541, \#3659) (\#3714)”; “Make the ship queue fetch the latest main before it checks a pull request (GST-2232) (\#4256)”; “ship accept: a lanes-only head policy gates the final tree; owner-granted test-lane exception (GST-1595) (\#4061)”; and 78 more (refs \#3541, \#3659, \#3714, \#4256, \#4061, …).
scripts/lib · medium confidence · unverified
9 commits (4 fixes) modifying ubicloud
A change to existing behaviour in ubicloud — 9 commits (4 fixs), 3 files: “fix(ubi-runner): an ambiguous create arms teardown only once this call's ssh key opens the VM (GST-1463) (\#4046)”; “v1.91.276.0 feat(preamble): tier 2+ skills load the shared preamble once per session (GSTA-96) (\#3434)”; “v1.93.6.0 feat: /plan-tune learns from ad-hoc question ids that carry a signal hint (\#3442) (\#3868)”; and 6 more (refs \#4046, \#3434, \#3442, \#3868, \#3839, …).
ubicloud · medium confidence · unverified
906 commits (439 fixes) modifying (repo-wide)
A change to existing behaviour in (repo-wide) — 906 commits (439 fixs), 32 files: “v1.92.48.0 fix(windows): review-scope summary runs gstack-review-read through Git Bash; jsonl-merge writes UTF-8 with LF (\#3541, \#3659) (\#3714)”; “v1.91.186.0 fix(redact): pii.cc skips compact YYYYMMDDHHMMSS stamps so autoplan approval imports stop failing 1 in 10 (GSTA-66) (\#3366)”; “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; and 903 more (refs \#3541, \#3659, \#3714, \#3366, \#2571, …).
(repo-wide) · low confidence · unverified
Autoplan review phases now use dual-voice (Claude + Codex) evaluation with consensus tables
The autoplan review process for the CEO, Design, DX, and Engineering phases has been updated to require a dual-voice evaluation: every phase now runs both a native Claude subagent review and an outside Codex review via Bash. The system generates a consensus table for each phase, comparing the two voices across specific dimensions (e.g., premises validity, design quality, developer experience, architecture soundness). If both voices agree on a change, it becomes a USER CHALLENGE; if they disagree on a valid strategic or technical reason, it is marked as a TASTE DECISION. The phases also include mandatory outputs like reality rows, error/rescue registries, and completion summaries, with specific degradation policies if one or both voices are unavailable.
autoplan/sections · high confidence
CSO security audit skill upgraded to v3 with evidence-based reporting and comprehensive repair workflows
The CSO skill has been updated to version 3.0.0, introducing a stricter evidence-based audit workflow that separates supported findings from labeled hypotheses. This update adds comprehensive mode capabilities, allowing the tool to generate isolated reproduction steps and up to three repair candidates when a qualified runtime catalog profile is available. The skill now enforces a private, isolated execution environment using a trusted launcher, ensuring that source inspection and scanner execution do not bypass redaction or affect the host. Reporting artifacts have been standardized to include stable finding IDs, explicit assurance labels (such as \self\_reported\ vs. \runtime\_tested\), and detailed coverage records, while legacy v2 report imports are supported in read-only mode.
cso · high confidence
Deterministic generation of app-owned bridge accessors
The iOS QA scripts now include a new tool (with a TypeScript fallback) that scans Swift source files for @Observable classes marked with @Snapshotable and automatically generates StateAccessor.swift files. This process is fully deterministic: it uses a composite cache key that includes the Swift version, tool revision, platform, and source content hash to ensure consistent output across different environments and builds, while strictly validating declarations to fail fast on invalid or duplicate keys. Also in this area: Unified state root for gstack data storage.
ios-qa/scripts · high confidence
Improved interactive element detection in browse snapshots
The browse snapshot command now automatically includes cursor-interactive scanning when the interactive flag is used, ensuring that dropdowns, popovers, and other non-ARIA interactive elements are captured. This change also adds specific detection for floating containers (like React portals) and removes the previous behavior that skipped elements with ARIA roles if they were missed by the accessibility tree, resulting in more complete and accurate snapshots for modern web applications.
browse · high confidence
Plan-eng-review skill receives major overhaul with scope gate and structured decision briefs
The plan-eng-review skill has been significantly updated to enforce a strict 'Scope gate' that auto-selects or prompts for the review target (branch diff, plan doc, or specific path) before any other work begins. The skill now includes a comprehensive 'AskUserQuestion Format' for structured decision briefs, complete with D-numbering, ELI10 explanations, completeness scores, and pros/cons, along with specific fallback behaviors for different session types (spawned, headless, interactive). Additionally, the skill's preamble has been enhanced to handle degraded modes, context recovery, and telemetry more robustly, and it now includes a 'Design Doc Check' and 'Scope Challenge' step to ensure thorough review preparation.
plan-eng-review · high confidence
Preamble instructions refactored into modular, context-aware generators
The system's preamble instructions have been restructured from monolithic inline scripts into a set of dedicated TypeScript generators (e.g., \generate-ask-user-format.ts\, \generate-completeness-section.ts\, \generate-preamble-bash.ts\). This change introduces context-aware rendering: the generated instructions now adapt to the specific skill (e.g., \plan-eng-review\), host environment (\gbrain\, \hermes\), and configuration (\explainLevel\). Key behavioral updates include a refined \AskUserQuestion\ protocol with specific handling for spawned vs. interactive sessions, a new \Completeness\ scoring principle for decision briefs, and a shift of heavy runtime bootstrapping (session bookkeeping, artifacts sync) to the \gstack-skill-start\ binary, leaving the preamble with only the interpretation rules the model needs. This results in a more modular, token-efficient, and robust instruction set that dynamically adjusts based on the current execution context. Also in this area: Windows shell bin fix and proactive skill defaults; Local evaluation runs now export EVALS\_RUN\_ID and confidence reviews use a single severity scale; New browse daemon route table and security hardening.
scripts/resolvers/preamble · high confidence
Renamed /checkpoint resume to /context-restore
The context-restore skill has been renamed from /checkpoint resume to /context-restore. This change was made because the /checkpoint command is now treated as a native rewind alias in current environments, which caused conflicts. The new skill loads the most recent saved state (preferring the current branch, falling back across branches) so users can pick up where they left off, even across Conductor workspace handoffs. It is designed to be used in conjunction with /context-save.
context-restore · high confidence
Renamed /checkpoint to /context-save to resolve command conflicts
The context-save skill has been renamed from /checkpoint to /context-save. This change resolves a conflict where the native /checkpoint command in Claude Code was shadowing the skill's functionality. The skill now triggers on 'save progress', 'save state', 'save my work', and 'context save', and it pairs with /context-restore to resume work in future sessions.
context-save · high confidence
Retro skill v2.0.0: new report structure, gbrain context, and robust date handling
The /retro skill has been upgraded to version 2.0.0, introducing a significantly more detailed retrospective report format that includes a tweetable summary, shipping velocity analysis, code quality signals, test health tracking, plan completion metrics, and a personal deep-dive for the user alongside a team breakdown. The skill now leverages gbrain to load prior retrospectives, recent timeline events, and learnings for trend analysis, while fixing date window calculations to use local midnight alignment instead of wall-clock time to prevent stale data issues. It also adds support for cross-project retrospectives, compare modes, and integrates with external context files like Greptile history and TODO lists to provide a more comprehensive view of engineering performance.
retro · high confidence
Ship workflow now detects base branch dynamically and supports Apple distribution
The ship skill has been refactored to replace the hardcoded 'main' branch assumption with dynamic base branch detection (Step 1), allowing it to operate correctly on any repository default or feature branch. Additionally, a new Apple target detection step (Step 0.9) has been added to handle App Store and TestFlight distribution workflows for Swift/Xcode projects, ensuring proper release pipelines are followed for iOS/macOS artifacts.
ship · high confidence
Telemetry schema expansion and security lockdown
The system now collects detailed usage and security telemetry, including new columns for tracking prompt injection attempts (attack domains, payload hashes, and verdicts) alongside standard skill run data. To protect this data, read access for anonymous users has been revoked via tightened Row Level Security policies, while write access is maintained through scoped, column-limited policies to support both legacy PostgREST clients and new edge functions. A cache table for community pulse aggregation has also been added to prevent performance degradation from repeated queries.
supabase/migrations · high confidence
Fixes
1 commit (1 fix) fixing design-review/templates
A fix in design-review/templates — 1 commit (1 fix), 1 file: “Add /design-review --build, an opt-in loop that checks and fixes new UI while you build it (\#4142)” (refs \#4142).
design-review/templates · medium confidence · unverified
29 commits (16 fixes) fixing scripts/ubicloud
A fix in scripts/ubicloud — 29 commits (16 fixs), 4 files: “fix(ubi-runner): an ambiguous create arms teardown only once this call's ssh key opens the VM (GST-1463) (\#4046)”; “v1.91.5.0 feat: balanced free-suite shards, test:ubicloud, and faster PR eval lane (\#2989)”; “v1.91.364.0 fix: pregate --policy-from head only adds lanes, and timeout\_s reaches preflight and finish (GST-419, GST-420) (\#3676)”; and 26 more (refs \#4046, \#2989, \#3676, \#3190, \#3697, …).
scripts/ubicloud · medium confidence · unverified
33 commits (20 fixes) fixing hosts/claude
A fix in hosts/claude — 33 commits (20 fixs), 23 files: “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; “v1.94.33.0 fix(ship-queue, hooks): selector reads the checkout under review; one SWITCH\_DEFAULT per hook switch; installs are not checks; timed agent deferral (GST-1088 follow-up) (\#3933)”; “v1.76.0.0 fix: ship doc-sync survives Conductor — spawned subagent sessions reachable (\#2733) (\#2741)”; and 30 more (refs \#2571, \#3933, \#2733, \#2741, \#3057, …).
hosts/claude · medium confidence · unverified
40 commits (22 fixes) fixing lib/pregate/checks
A fix in lib/pregate/checks — 40 commits (22 fixs), 8 files: “v1.91.357.0 fix(pregate): literals skips old rows of a version-history table the branch extended (GST-344) (\#3669)”; “v1.91.364.0 fix: pregate --policy-from head only adds lanes, and timeout\_s reaches preflight and finish (GST-419, GST-420) (\#3676)”; “v1.91.107.0 fix: the GBRA-40 field list (one Bun for install and check, \rec\ gate approval, pregate never silent) (\#3190)”; and 37 more (refs \#3669, \#3676, \#3190, \#3782, \#3579, …).
lib/pregate/checks · medium confidence · unverified
Fixes command parsing and decision envelope in the /careful safety hook
The /careful bash hook now uses a shared JSON extractor (hook-extract.sh) to correctly parse command arguments containing escaped quotes, preventing destructive commands from being silently truncated and allowed. It also ensures safety decisions are nested under hookSpecificOutput so Claude Code does not ignore them, and fails closed when the input cannot be parsed or helpers are missing. Also in this area: Prevents false positives in destructive checks by masking git commit messages; Unified state root for gstack data storage.
careful/bin · high confidence
Migration scripts for gstack upgrades v0.15.2.0 through v1.86.0.0
This update adds a suite of 16 migration scripts to the \gstack-upgrade/migrations\ directory, covering versions from v0.15.2.0 to v1.86.0.0. These scripts automate the transition for existing users by fixing skill directory structures for unprefixed discovery (v0.15.2.0), merging per-project resource logs into the builder profile (v0.16.2.0), introducing a one-time prompt for the new V1 writing style (v1.0.0.0), and removing the stale \/checkpoint\ skill to prevent shadowing Claude Code's native \/rewind\ alias (v1.1.3.0). Further migrations wire existing brain-sync repositories as gbrain federated sources (v1.17.0.0), rename \gstack-brain-\\ to \gstack-artifacts-\\ including repo and config updates (v1.27.0.0), and notify users about split-engine gbrain with local PGLite code search (v1.37.0.0). Additional scripts patch allowlists and privacy maps for new artifact patterns (v1.38.1.0, v1.40.0.0), register Conductor AskUserQuestion hooks (v1.58.0.0), repair poisoned Chrome-for-Testing bundles on macOS (v1.65.0.0), restore tracked files dirtied by legacy in-place gbrain renders (v1.67.0.0), migrate feature-discovery acknowledgements to \GSTACK\_HOME\ (v1.78.0.0), and update \/claude\ to \/claude-code\ on non-Claude hosts (v1.86.0.0).
gstack-upgrade/migrations · high confidence
New telemetry and update-check edge functions with security and reliability improvements
Added three new Supabase Edge Functions: \telemetry-ingest\ for validating and storing telemetry events, \community-pulse\ for aggregating community statistics, and \update-check\ for logging install pings. The \telemetry-ingest\ function improves security by using the Supabase anon key with Row Level Security (RLS) instead of the over-privileged service role key, and enforces payload size and batch limits. The \community-pulse\ function fixes a reliability issue where database errors previously returned fake zero counts; it now explicitly checks for query errors and serves stale cache or a 503 error instead. It also introduces k-anonymity filtering for security event data to prevent re-identification of individual users' attack logs. The \update-check\ function ensures the current version is always returned to the client, even if the background logging to the database fails.
supabase/functions · high confidence
Reliable mode-handoff message display in Claude Code sessions
A new bash shim and TypeScript hook in plan-ceo-review/bin ensure that the 'mode handoff' status line (e.g., 'Auto-decided review mode → HOLD SCOPE') is reliably displayed to the user as a system message. Previously, Claude Code's output collapsing and model paraphrasing often hid this critical context. The shim handles Windows/MSYS path conversion issues by passing NUL or /dev/null as the Bun config, and the hook specifically intercepts the output of the \gstack-ceo-mode-handoff\ command to extract and present the exact status line, ensuring users always see the current review mode regardless of subsequent model behavior.
plan-ceo-review/bin · high confidence
Supabase RLS verification and configuration scripts
Added \config.sh\ to store public Supabase connection details and \verify-rls.sh\ to validate Row Level Security policies. The verification script ensures that read and update operations are denied via the anonymous key while insert operations remain allowed for backward compatibility, confirming the security posture of the deployed RLS rules.
supabase · high confidence
Test coverage
1 commit adding/updating tests in (repo-wide); 307 commits adding/updating tests in test/fixtures; 341 commits adding/updating tests in test/helpers; 48 commits adding/updating tests in spec; 8 commits adding/updating tests in crew/test; 951 commits adding/updating tests in test; Added comprehensive test coverage for the make-pdf module; Added end-to-end gate tests for make-pdf output quality and format fidelity; Added provider adapter tests for Claude, Gemini, and GPT; Added test coverage for browse module security, activity, and browser management; Added test coverage for quality gate secret-sink invariant; Expanded test fixtures for browser automation and security validation.
Dependencies
750 commits updating dependencies (9 manifests)
A dependency / build maintenance change in (dependencies) — 750 commits (390 fixs), 9 files: “v1.92.48.0 fix(windows): review-scope summary runs gstack-review-read through Git Bash; jsonl-merge writes UTF-8 with LF (\#3541, \#3659) (\#3714)”; “v1.91.186.0 fix(redact): pii.cc skips compact YYYYMMDDHHMMSS stamps so autoplan approval imports stop failing 1 in 10 (GSTA-66) (\#3366)”; “v1.64.0.0 fix wave: full tracker audit — 90 fixes, 52 issues closed, \~50 community PRs absorbed (\#2571)”; and 747 more (refs \#3541, \#3659, \#3714, \#3366, \#2571, …).
(dependencies) · high confidence · unverified
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
How this codebase got here
Score
- CAI 43 → 41 (-1.9)
- Rubric changed (rubric-2026.10.5 → rubric-2026.11.4) — scores are not directly comparable.
Lenses
- Code Health 52 → 52 (+0.2)
- Architecture 75 → 60 (-15.2)
- Maturity 84 → 87 (+3.2)
- Readiness 55 → 55 (-0.1)
- Security 61 → 52 (-9.2)
- Accessibility 26 → 25 (-1.7)
- Performance 60 → 60 (+0.0)
Resolved (363)
- BrowserManager.launch (cognitive 16) (browse/src/browser-manager.ts)
- ClassTooLong: BrowserManager (browse/src/browser-manager.ts)
- FileTooLong: lib/paid-plan.ts (scripts/lib/paid-plan.ts)
- FileTooLong: lib/paid-report.ts (scripts/lib/paid-report.ts)
- FileTooLong: scripts/cso-eval-producer.ts (scripts/cso-eval-producer.ts)
- FileTooLong: scripts/cso-eval.ts (scripts/cso-eval.ts)
- FileTooLong: scripts/gen-skill-docs.ts (scripts/gen-skill-docs.ts)
- FileTooLong: scripts/ship-measure.ts (scripts/ship-measure.ts)
- FunctionTooLong: cso-eval-producer.runProducerCell (scripts/cso-eval-producer.ts)
- FunctionTooLong: gate-reuse.findReusableGate (scripts/gate-reuse.ts)
- FunctionTooLong: paid-plan.buildRunManifest (scripts/lib/paid-plan.ts)
- FunctionTooLong: paid-plan.parseRunManifest (scripts/lib/paid-plan.ts)
- FunctionTooLong: paid-report.runPaidReport (scripts/lib/paid-report.ts)
- Hotspot: .github/scripts/qualify-dia-macos.ts (.github/scripts/qualify-dia-macos.ts)
- Hotspot: .github/scripts/run-dia-native-qualification.ts (.github/scripts/run-dia-native-qualification.ts)
- Hotspot: design/src/generate.ts (design/src/generate.ts)
- Hotspot: extension/background.js (extension/background.js)
- Hotspot: ios-qa/daemon/src/tunnel-bootstrap.ts (ios-qa/daemon/src/tunnel-bootstrap.ts)
- Hotspot: lib/ship-queue-train-ready.ts (lib/ship-queue-train-ready.ts)
- Hotspot: lib/ship-queue-train-verify.ts (lib/ship-queue-train-verify.ts)
- …and 343 more
New (975)
- Change coupling: capy-review-pass.js ↔ review-pass.ts (lib/capy-review-pass.js)
- ClassTooLong: TabScope (browse/src/browser-manager.ts)
- FileTooLong: lib/design-catalog.ts (lib/design-catalog.ts)
- FileTooLong: lib/eval-plan.ts (lib/eval-plan.ts)
- FileTooLong: lib/harness-register.ts (lib/harness-register.ts)
- FileTooLong: lib/restamp.ts (lib/restamp.ts)
- FileTooLong: lib/ship-policy.ts (lib/ship-policy.ts)
- FileTooLong: lib/ship-queue-train.ts (crew/lib/ship-queue-train.ts)
- FileTooLong: lib/ship-release.ts (lib/ship-release.ts)
- FileTooLong: src/diagram-prepass.ts (make-pdf/src/diagram-prepass.ts)
- Fixed-sleep synchronisation: a Stop that fires before the CLI flushes the load itself still checks the run once the final message lands (test/skill-contract-stop-hook.test.ts)
- Fixed-sleep synchronisation: a Stop that fires before the CLI flushes the load itself still judges the run once the final message lands (test/ceo-review-gate.test.ts)
- Fixed-sleep synchronisation: a Stop that fires before the CLI flushes the turn waits for the final message instead of judging a stale transcript (test/autoplan-stop-gate.test.ts)
- Fixed-sleep synchronisation: a Stop that fires before the CLI flushes the turn waits for the final message: a review written outside the repo is released, not held (test/ceo-review-gate.test.ts)
- Fixed-sleep synchronisation: a Stop that fires before the CLI flushes the turn waits for the final message: a run that reran its check is met, not blocked (test/skill-contract-stop-hook.test.ts)
- Fixed-sleep synchronisation: a call in flight when the process died is charged its reservation on the next run, which dispatches a fresh attempt (test/autoplan-run-api.test.ts)
- Fixed-sleep synchronisation: a cap changed during the call admits the retry against the new cap and is not written back over (test/autoplan-shadow.test.ts)
- Fixed-sleep synchronisation: the bin dies on SIGTERM at once (no JS handler swallows it) and the next watcher takes over its lock (test/pr-wait.test.ts)
- FunctionTooLong: claude-code-migration.migrateClaudeCodeSkills (lib/claude-code-migration.ts)
- FunctionTooLong: gate-reuse.findReusableGate (crew/scripts/gate-reuse.ts)
- …and 955 more
Changes since last survey
- 300 commits — 180 feature/other, 120 fixes
By area
- (root) — 69 commits
- test/fixtures — 27 commits
- ship/sections — 15 commits
- .github/workflows — 10 commits
- test/helpers — 10 commits
- bin/gstack-review-dispatch — 8 commits
- bin/gstack-ship-receipt — 7 commits
- capy/skill — 7 commits
- bin/gstack-doctor-components.sh — 6 commits
- browse/src — 6 commits
- hosts/claude — 6 commits
- lib/pregate — 6 commits
- bin/gstack-review-plan — 4 commits
- browse/test — 4 commits
- .gstack/ship-policy.json — 3 commits
- bin/gstack-artifacts-init — 3 commits
- bin/gstack-uninstall — 3 commits
- captain/templates — 3 commits
- crew/lib — 3 commits
- .gstack/pregate.json — 2 commits
Notable commits
- fix: Add /design-review --build, an opt-in loop that checks and fixes new UI while you build it (#4142)
- fix: Capy session: defer the upgrade while a /ship or review run is open; upgrade-path fixes (GST-1677, GST-1658) (#4095)
- fix: Fix five small browse and question-hook bugs (#4192)
- fix: Fix seven test and eval checks that misread correct work or raced their neighbours (GST-2131) (#4223)
- fix: Fix ship-queue ready tests after the pre-merge check got stricter (GST-2282) (#4271)
- fix: Fix six test and eval checks that misread correct work or raced their own fixtures (GST-2338) (#4284)
- fix: Fix three flaky test causes found in the 2026-11-08 harvest (GST-1912) (#4167)
- fix: Keep crew threads on their PR to the end and cap finished PRs at 10 before new fix threads (GST-2192) (#4240)
- fix: Teach crews to keep work owned and treat outside PRs as bug reports (GST-2077) (#4204)
- fix: fix(autoplan): bind rec_hash to its run and plan; keep legacy gates answerable (GST-1819) (#4137)
- fix: fix(autoplan): export --public drops owner replies and redacts quoted owner words (GST-1243) (#3967)
- fix: fix(browse): touch points wait for a rendered frame, so a tap right after goto lands where it was aimed (#3956)
- fix: fix(capy-install): read project pins with an isolated Bun (GST-1747) (#4115)
- fix: fix(ci-carry): a per_file remainder file that skips here too is not owed (GST-1314) (#3985)
- fix: fix(design): exits ease out, never ease in (GST-1240) (#3963)
- fix: fix(doctor): parse the launch probe under macOS /bin/bash 3.2 (#3601) (#3753)
- fix: fix(evals): repair the two red Codex census cases (cso-diff-mode fixture, pair-agent divider wording) (#3800)
- fix: fix(evals): screen build-image keys the CI image on the merge commit it tests (#3962)
- fix: fix(evals): the Codex removed-rule arm really removes the rule from installed skills (GST-840) (#3816)
- fix: fix(flaky): brain-sync auth check ignores the remote URL, browse binds its port after launch, three paid-eval clarity repairs (GST-1438) (#4035)
- …and 280 more
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
Survey your own repository
garrytan/gstack was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.
About this page
- The score is its most recent published measurement, taken on 14 November 2026 at a pinned commit. It is not a live figure and does not change until the project is measured again.
- Measured at commit 3b650356f3d711caca98713475dd6085b15c5996 — the exact code this score is about.
- Scored under rubric-2026.11.4 — the same rubric and the same method as every other entry in this index.
- Measured by Canine Code Analyzer (cca.canine.dev) using cca measure-2026.11.65+preprod-e1a3b69fc8ee.