Lightweight URL/tool triage that sits between \"ignore\" and /url-dump. Evaluates whether a URL or tool is worth saving or skipping — checking existing vault coverage, assessing relevance to the user's profile and interests, and recommending a clear next action.
Check agent_mode in 00-inbox/MY-PROFILE.md frontmatter:
agent_mode: team — delegate vault scanning and web fetching to parallel sub-agents (one for vault search, one for content fetch/analysis). Combine results for recommendation.agent_mode: solo (default) — handle all scanning and analysis directly in the conversation. No delegation.Before executing, check for user profile:
\n00-inbox/MY-PROFILE.md and 00-inbox/MY-INTERESTS.md in the vaultWelcome to COG! Scout works best with a profile for relevance matching.\n\nWould you like to run /onboarding first, or should I evaluate with general criteria?\nMY-PROFILE.md for active projects and roleMY-INTERESTS.md for topic areas00-inbox/MY-INTEGRATIONS.md for active integrations (check if web-fetch and web-search are available)/url-dumpScout evaluates (\"should I save this?\"). URL-dump saves (\"save this now\").
\n/url-dump with pre-filled category/url-dump directlyAccept one or more of:
\nPrompt (if no input provided):
\nWhat URL(s) or tool(s) would you like me to evaluate?\n(You can paste URLs, tool names, or a mix)\nBatch mode: Multiple URLs/names in one invocation are processed together with a summary table at the end.
\nFor each URL or tool name, search the entire vault for existing coverage.
\nSearch strategy:
\ngithub.com/owner/repo → search for repo name)url: fields) and inline linksIf found:
\n\uD83D\uDD0D Existing coverage found for [name]:\n- [file path] — saved [date], category: [category]\n- [file path] — mentioned in [context]\n\nWant me to check if an update is needed, or skip this one?\nIf URL provided and web-fetch is active:
\nIf tool name provided (no URL):
\nContent type detection:
\nScore relevance against user context:
\nProfile Match (from MY-PROFILE.md):
\nInterest Match (from MY-INTERESTS.md):
\nQuality Signals:
\nBased on analysis, recommend one of two actions:
\n✅ SAVE — [Title/Name]\nCategory: [suggested category for url-dump]\nRelevance: [High/Medium] — [why it matters]\nProjects: [affected project(s) if any]\n\nShall I hand off to /url-dump to save it?\n⏭️ SKIP — [Title/Name]\nReason: [clear explanation — wrong stack, low quality, already covered, irrelevant to interests]\nWhen processing multiple URLs/tools, end with a summary table:
\n## Scout Summary\n\n| # | Item | Verdict | Reason |\n|---|------|---------|--------|\n| 1 | [Name 1] | ✅ Save | [brief reason] |\n| 2 | [Name 2] | ⏭️ Skip | [brief reason] |\n\n**Actions:**\n- [X] items ready to save via /url-dump\nBased on user confirmation:
\n/url-dump with pre-filled category suggestion| Scenario | \nBehavior | \n
|---|---|
| web-fetch unavailable | \nEvaluate based on URL structure, domain reputation, and vault search only. Note that content wasn't fetched. | \n
| web-search unavailable | \nFor tool-name inputs (no URL), ask the user for a direct URL instead. For URL inputs, proceed normally — web-search is not needed. | \n
| No user profile | \nEvaluate with general quality/relevance criteria, skip personalized relevance scoring | \n
| URL is paywalled | \nNote limitation, evaluate based on available preview and metadata | \n
| Tool not found via search | \nAsk user for more context or a direct URL | \n
/url-dump with pre-filled category/daily-brief may surface new tools/services → user can run /scout to evaluate/auto-research may discover tools during research → scout can triage them/url-dump when savingScout embodies COG's \"evaluate before you accumulate\" principle:
\n