How do I install the Experience Recall skill?

Run npx skillmds@latest add mace-enterprise/experience-recall in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Experience Recall safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Experience Recall?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Experience Recall free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Experience Recall?

Mace-Enterprise (@mace-enterprise) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Experience Recall by mace-enterprise

Name: Experience Recall
Author: Mace-Enterprise

Experience Recall

Look up prior development learnings in the knowledge wiki before starting non-trivial work, so past patterns, gotchas, and verifier/environment recipes are reused instead of rediscovered. Use before designing or implementing a change, or when the user asks "have we solved this before / what do we know about X".

File contents

Experience recall

Query the cross-project experience library before acting, so you reuse what the team already learned.

Preferred: the recall MCP

If the knowledge-recall MCP is registered, use it first — it's hybrid (semantic + keyword) recall over the whole wiki and finds by concept, not just exact words:

search_notes(query, mode="hybrid", limit=8) → ranked pages with snippets.
read_note(path) → the full page.

Then read the top 1–3 hits and cite them. If the MCP is not available, fall back to the index-first + grep procedure below.

Where to look (fallback when the MCP isn't registered)

The wiki at <WIKI_DIR> (canonical protocol: WIKI_PROTOCOL.md):

Wiki/index.md — the table of contents. Read this first; scan the relevant layers for the task.
Wiki/experience/<category>/*.md — verified learnings (patterns / gotchas / verifiers / environments / methodology).
Wiki/knowledge/*.md — source-distilled knowledge.
Wiki/adr/*.md — past decisions (why we chose X); Wiki/guardrails/*.md — rules that constrain how to act; Wiki/journal/sessions/*.md — prior discussion/context (may be unverified — treat as leads, not facts).

Procedure

Read Wiki/index.md. Identify candidate pages by title/description and by the current task's stack (e.g. dotnet/wpf) and category (a bug → gotchas; "how do I run/test this" → environments/verifiers).
Grep the experience folders for keywords from the task if the index is thin.
Read the 1–3 most relevant pages in full.
Apply what you find. Cite the page you used (e.g. "per [[ib-eventaggregator-ui-thread]]") so the user sees the library paying off.
If nothing relevant exists, say so briefly and proceed — and remember this gap may become a new experience page once the work is verified (see /learn).

Keep it lightweight: this is a read-only lookup, not a full audit.

---name: experience-recalldescription: Look up prior development learnings in the knowledge wiki before starting non-trivial work, so past patterns, gotchas, and verifier/environment recipes are reused instead of rediscovered. Use before designing or implementing a change, or when the user asks "have we solved this before / what do we know about X".---# Experience recallQuery the cross-project experience library before acting, so you reuse what theteam already learned.## Preferred: the recall MCPIf the **`knowledge-recall`** MCP is registered, use it first — it's hybrid(semantic + keyword) recall over the whole wiki and finds by *concept*, not justexact words:- `search_notes(query, mode="hybrid", limit=8)` → ranked pages with snippets.- `read_note(path)` → the full page.Then read the top 1–3 hits and cite them. If the MCP is **not** available, fallback to the index-first + grep procedure below.## Where to look (fallback when the MCP isn't registered)The wiki at `<WIKI_DIR>` (canonical protocol:`WIKI_PROTOCOL.md`):- `Wiki/index.md` — the table of contents. Read this **first**; scan the relevant layers for the task.- `Wiki/experience/<category>/*.md` — verified learnings (patterns / gotchas / verifiers / environments / methodology).- `Wiki/knowledge/*.md` — source-distilled knowledge.- `Wiki/adr/*.md` — past decisions (why we chose X); `Wiki/guardrails/*.md` — rules that constrain how to act; `Wiki/journal/sessions/*.md` — prior discussion/context (may be unverified — treat as leads, not facts).## Procedure1. Read `Wiki/index.md`. Identify candidate pages by title/description and by the current task's stack (e.g. dotnet/wpf) and category (a bug → gotchas; "how do I run/test this" → environments/verifiers).2. Grep the experience folders for keywords from the task if the index is thin.3. Read the 1–3 most relevant pages in full.4. Apply what you find. **Cite the page** you used (e.g. "per [[ib-eventaggregator-ui-thread]]") so the user sees the library paying off.5. If nothing relevant exists, say so briefly and proceed — and remember this gap may become a new experience page once the work is verified (see `/learn`).Keep it lightweight: this is a read-only lookup, not a full audit.

Frequently asked questions