# Share Learning > Promote a team-relevant learning to the shared team-knowledge repo after deduping against existing notes. Triggers "share this", "add to the knowledge base", or a gotcha, decision, or convention worth team-wide awareness. - Skill: `darkroomengineering/share-learning` (Agent Skill) - Install (CLI): `npx skillmds@latest add darkroomengineering/share-learning` - Raw SKILL.md: https://api.skillmd.com/api/skills/darkroomengineering/share-learning/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: darkroomengineering (https://skillmd.com/u/darkroomengineering) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/darkroomengineering/share-learning --- # share-learning Promote a single learning to the team's shared knowledge repo (`darkroomengineering/team-knowledge`) — the "public corpus" tier of the knowledge system (see `docs/knowledge-system.md`). Local, personal knowledge stays in auto-memory; this skill is only for things another teammate's agent would benefit from knowing. ## When to use Use when a learning meets the shared-tier bar from `AGENTS.md` (Knowledge Routing): an architecture decision the team must follow, a library gotcha that affects everyone, a convention, an incident postmortem, or a reusable pattern. If it is a personal preference, local project state, or an external pointer, let auto-memory handle it instead — do NOT post it. ## Inputs Invoked as `/share-learning ""` where `` is one of: `decision`, `convention`, `gotcha`, `incident`, `pattern`. If invoked without arguments, infer the most likely `kind` and a concise `text` from the recent conversation, then show the user what you intend to post and confirm before posting. Before starting, run `gh auth status` and confirm the authenticated account can read `darkroomengineering/team-knowledge` (or `$KNOWLEDGE_REPO`) with `gh api repos/${KNOWLEDGE_REPO:-darkroomengineering/team-knowledge} --jq .full_name`. Stop with the failed prerequisite if either check fails. Do not wait until the write step to surface missing authentication or repository access. ## Steps 1. **Resolve the repo.** Read `$KNOWLEDGE_REPO` from the environment; default is `darkroomengineering/team-knowledge`: ```bash KNOWLEDGE_REPO="${KNOWLEDGE_REPO:-darkroomengineering/team-knowledge}" ``` 2. **Dedup against the index (required).** Fetch the current index: ```bash gh api repos/$KNOWLEDGE_REPO/contents/INDEX.md --jq .content | base64 -d ``` Scan the note names and titles in the index for an entry that already captures this learning (semantic near-duplicate, not just exact match). If you find one: - Show the user the existing note name and its summary line. - Ask whether to **skip** (already covered), **post anyway** (genuinely distinct), or **revise** your proposed entry to complement it. Only continue to step 3 once the user has chosen, or when there is clearly no duplicate. 3. **Post.** Derive a `name` (kebab-case slug from the essence of the learning). Assemble the note: - Frontmatter: `name` = the slug; `kind` from the argument; `added-by` from `gh api user --jq .login` (fall back to `git config user.name` if that fails); `tags` optional; `supersedes` only when this note replaces an existing one. - Body: what happened + why it matters + how to apply it. One learning per note, atomic and self-contained. If creating a new note: ```bash NOTE="--- name: kind: tags: [, ] added-by: --- " gh api -X PUT repos/$KNOWLEDGE_REPO/contents/.md \ -f message="knowledge: add " \ -f content="$(printf '%s' "$NOTE" | base64)" ``` If updating an existing note, first GET its current `sha`: ```bash SHA=$(gh api repos/$KNOWLEDGE_REPO/contents/.md --jq .sha) gh api -X PUT repos/$KNOWLEDGE_REPO/contents/.md \ -f message="knowledge: update " \ -f content="$(printf '%s' "$NOTE" | base64)" \ -f sha="$SHA" ``` 4. **Report.** Surface the blob URL to the user: `https://github.com/$KNOWLEDGE_REPO/blob/main/.md` ## Notes - This skill posts to a shared, team-visible repo — treat it like publishing. Never post secrets, credentials, or anything from `.env`. When unsure whether something is team-relevant, ask the user rather than over-sharing. - The dedup step is what makes this more than a `gh` wrapper: you are exercising judgment about whether the corpus already knows this.