Capture and classify learnings into policy rules or insights with qmd dedup.
npx skills add https://github.com/indigoai-us/hq --skill learn-indigoai-us
Please help me install this Agent Skill. Skill: learn Source: https://github.com/indigoai-us/hq/tree/main/template/.claude/skills/learn Command: npx skills add https://github.com/indigoai-us/hq --skill learn-indigoai-us
Capture and classify learnings from tasks, sessions, or manual input, then route them into structured policy files (rules) or insight files. It automatically deduplicates using qmd similarity checks and rebuilds the policy digest after changes. The skill is callable programmatically or from common HQ entry points such as /execute-task, /run-project, /handoff, and /checkpoint, with a --hard flag to enforce hard rules.
Provide a learning description to classify it into a policy rule or an insight and trigger the dedup and digest workflow.
High-intent search queries and answers about installing and using this skill.
To convert learnings into policy rules, the system captures and classifies inputs from tasks, sessions, or manual submissions, then routes them into structured policy files or insight files for governance and storage.
The best way to prevent duplicate knowledge management entries is to use qmd similarity checks with a fallback grep-based scan. This deduplication workflow triggers automatically after capturing new learnings to ensure unique records.
A policy digest updates automatically by rebuilding after changes are made to the rule files. This ensures that newly captured learnings and policies are loaded and available for enforcement in future sessions.
Yes, you can programmatically invoke learning ingestion during task execution using CLI-based entry points like /execute-task, /run-project, /handoff, and /checkpoint to integrate capture into automated workflows.
You should use hard enforcement for governance policies when strict compliance is required. Applying the --hard flag designates captured learnings as hard-enforcement rules rather than flexible insights during the classification process.
SilentFlowerLoad workflow guides, git status, and project context to initialize AI development sessions.
raopriyamAutomate Nx tasks across projects for build, test, lint, or serve commands.
w-winterSearch indexed text corpora with semantic ranking and structured queries.
muhammadhamad1104Manage and retrieve project learnings stored in a JSONL file.
