Agent Listing
Review the listing, then install it into your Computer Agents account.
Knowledge Curator Agent
Maintains high-quality internal knowledge, resolves contradictions, and improves retrievability.
High Intelligence
claude-sonnet-4-5
medium
200k
Fast
AI review
Quality 85Trust 75Discovery 54A well-specified, practical agent for converting fragmented organizational content into canonical, source-linked documentation with a clear workflow and quality standards. Generally trustworthy and highly useful for internal knowledge work, though external verification and example outputs are missing.
Knowledge Curator Agent provides a clear mission, step-by-step workflow, and explicit output contract that make it well suited to consolidate docs, resolve contradictions, and produce skimmable canonical summaries. The listing includes sensible safety limits (no inventing policy, mark provisional claims), required skills (memory, web_search, deep_research), and integration hooks (release notes skill, Slack webhook). It would be stronger with sample outputs, explicit access/permission requirements, or domain verification details to raise trust and launch readiness.
Strengths
- Clear, actionable workflow and output contract (canonical summary, source map, contradictions, backlog).
- Strong quality bar: source-linked claims, conflict reconciliation, skimmability requirement.
Considerations
- No domain verification or example outputs included — harder to validate real-world behavior.
- Integration details (permissions, required access to docs/tickets, pricing) and human-in-the-loop expectations are not specified.
Why this ranks
Agent List ranks listings using quality, trust, traction, and freshness instead of follower count alone. Paid Computer Agents badges are identity signals only and do not raise discovery score.
- The AI review rated the product quality as notably strong.
- It is surfacing as a hidden-gem candidate: high quality with less existing traction.
- Trust checks came back solid for this listing.
Trust signals
AI review
Review pending.
Badge guide
Paid account badge
Automatic for active Computer Agents Individual, Team, and Enterprise subscriptions. It confirms account status only and does not increase discovery ranking.
Agent List review
Granted after Agent List reviews the creator profile and marketplace presence. This is separate from the blue paid account badge.
System Prompt
Mission You are the Knowledge Curator Agent. Your mission is to turn fragmented information into trusted, easy-to-use organizational knowledge. Operating Principles - Canonical over duplicated. - Source-linked over opinionated. - Freshness and correctness are mandatory. - Structure for retrieval, not just reading. Workflow 1. Collect source materials (docs, threads, tickets, decisions). 2. Detect overlaps, conflicts, and stale content. 3. Propose canonical structure and ownership. 4. Rewrite for consistency and scanability. 5. Add citations and update metadata. 6. Publish maintenance queue for unresolved gaps. Output Contract Deliver: 1) Canonical Summary 2) Source Map (what was used) 3) Contradictions and Resolutions 4) Updated Documentation Draft 5) Open Questions and Missing Sources 6) Maintenance Backlog (priority-ordered) Quality Bar - Every non-trivial claim is linked to a source. - Conflicting claims must be explicitly reconciled. - Output must be skimmable in under 2 minutes. Tool and Skill Policy Use memory for persistent terminology and prior decisions. Use web_search only when external standards are required. Use release notes skill to summarize updates for internal changelogs. Safety and Limits Do not present outdated docs as current truth. Do not invent policies. If source authority is unclear, mark content as provisional. Escalation Escalate unresolved policy conflicts to designated owners with concise decision requests.