Skip to content
CAI
Software that uses CAICheck a score

anthropics/knowledge-work-plugins

54.9

Weak · 18 September 2026

8k

lines of production code

Python

primary language

1

measurement over time

CAI band scale
CAI lens gauges

What this system is

This system is a modular plugin architecture that extends functionality through specialized, partner-built, and internal skill sets. It currently supports sales prospecting and brand voice enforcement via external integrations, while providing structured behavioral guidelines for small business operations. The platform also manages productivity workflows and previously handled clinical trial protocols, though the latter capability has been removed.

Features

New partner plugins: Apollo and Brand Voice

Two new partner-built plugins are now available. The Apollo plugin (v0.1.0) adds three sales skills—lead enrichment, ICP-based prospecting, and sequence loading—via an HTTP MCP server at https://mcp.apollo.io/mcp. The Brand Voice plugin (v1.1.0) by Tribe AI discovers brand materials and enforces voice across enterprise platforms, connecting to Notion, Atlassian Confluence, Box, Figma, Gong, and Granola via HTTP MCP servers (e.g., https://mcp.notion.com/mcp, https://mcp.atlassian.com/v1/mcp, https://mcp.box.com, https://mcp.figma.com/mcp, https://mcp.gong.io/mcp, https://mcp.granola.ai/mcp) and leveraging native integrations for Google Drive and Slack.

partner-built · high confidence

Removals

Removal of clinical trial protocol generation skill

The clinical trial protocol generation capability within the bio-research plugin has been removed. This change deletes the main orchestrator skill (\clinical-trial-protocol/SKILL.md\), all associated reference sub-skills (covering intervention initialization, regulatory research, protocol foundation, intervention details, and operations), the FDA clinical protocol template, and the statistical sample size calculator script. Users will no longer be able to generate or manage clinical trial protocols for medical devices or drugs through this plugin.

bio-research/skills · high confidence

Behavioural changes

Improved dashboard security and restructured productivity skills

The productivity dashboard now escapes file-derived content before rendering it via innerHTML, preventing potential cross-site scripting issues when displaying task titles, notes, and memory file names. Additionally, the system has migrated from a 'commands' to a 'skills' structure, renaming and relocating skill definitions (such as start, update, and memory-management) to the new skills directory, with updated descriptions and path references to reflect the new organization.

productivity/skills · high confidence

New shared behavioral guidelines for the Small Business plugin

The small-business/shared area now includes a suite of new reference documents that define plugin-wide behavioral rules and connector-specific implementation details. These files establish how skills should handle data integrity (distinguishing absent data from zero), manage cross-skill data seams, enforce connector neutrality, and format outputs (including currency, locale, and document preferences). They also provide specific traps and correct API call shapes for connectors like Gmail, Gusto, QuickBooks, HubSpot, and Salesforce, ensuring consistent and accurate behavior across the plugin.

small-business/shared · high confidence

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

How this codebase got here

Baseline

  • First survey — no prior run to compare against. CAI 55.

Lenses

  • Code Health 83
  • Architecture 99
  • Maturity 78
  • Readiness 32
  • Security 100
  • Accessibility 65

Changes since last survey

  • 300 commits — 300 feature/other, 0 fixes

By area

  • .claude-plugin/marketplace.json — 176 commits
  • (repo) — 121 commits
  • sales/.claude-plugin — 1 commit
  • sales/skills — 1 commit
  • small-business/skills — 1 commit

Notable commits

  • change: Add BlackRock Advisor Center plugin (#1125)
  • change: Add Chronograph GP and LP plugins (chronograph-gp, chronograph-lp) (rebased on main) (#1179)
  • change: Add Claude for Financial Advisors plugin (#1130)
  • change: Add Informatica plugin (informatica-for-claude-platform) (#1146)
  • change: Add Intuit QuickBooks plugin (intuit-quickbooks) (#1159)
  • change: Add Leadfeeder plugin (leadfeeder) (#1162)
  • change: Add Wingspan plugin (wingspan) (#1196)
  • change: Add Zocks Advisor Intelligence plugin (zocks-advisor) (#1158)
  • change: Add activecampaign plugin (#1016)
  • change: Add gc-ai plugin
  • change: Add hubspot-sales plugin
  • change: Add incident.io plugin (incident-io) (#1182)
  • change: Add maven-bio plugin (#1109)
  • change: Add netsuite-ai-companion and netsuite-finance-analyst plugins (#1108)
  • change: Add noibu plugin (#806)
  • change: Add qodo-standards plugin (Qodo, packages/qodo-standards @ v1.0.10) (#1065)
  • change: Add security-guidance plugin
  • change: Add unity plugin (#965)
  • change: Add unstructured-foundation plugin
  • change: Claude for Small Business launch (#1144)
  • …and 280 more

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

Survey your own repository

anthropics/knowledge-work-plugins 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 18 September 2026 at a pinned commit. It is not a live figure and does not change until the project is measured again.
  • Measured at commit 67e216dbe5621470581dd3d944df22d1cc8231ce — the exact code this score is about.
  • Scored under rubric-2026.09.15 — the same rubric and the same method as every other entry in this index.
  • Measured by watchdog.canine.dev using codehealth-analyzer preprod-5d04157a340d.