temper-pass / rules
Cursor rule0 starsChanged 47 days agoMITView on GitHub
Before substantial feature work, run ForgeTrail lessons + anti-patterns (MCP) and fold findings into the plan
Where files are headingNew Cursor rule files have more rules about planning and tracking tasks, and fewer about code style.
Of the Cursor rule files we read, the 4,157 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository8 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .forgetrail/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .forgetrail/CLAUDE.mdCLAUDE.md
- .cursor/rules/commit-messages.mdcCursor rule
- .cursor/rules/forgetrail-lessons-mcp.mdcCursor rule
- .cursor/rules/forgetrail-phase-status.mdcCursor rule
- .cursor/rules/git-user-commits.mdcCursor rule
The file
We split this forgetrail-lessons-gate.mdc into the rules it gives an agent
6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
Run getAntiPatterns once and searchLessons (one or more queries) when any of these is true:
- Very specific: line 23
If the client exposes tool schemas, use them for searchLessons (query + optional limit) and getAntiPatterns (no arguments).
- Very specific: line 24
Call getAntiPatterns (no args).
- Very specific: line 25
Call searchLessons with 2–4 short keyword queries derived from the work (e.g.
- Distinctive: line 26
Fold in before coding: Add 3–6 bullet points to your plan or first implementation message—what you will do or avoid (e.g.
- Distinctive: line 30
Do not mention MCP, tool names, or “lessons gate” in UI copy or user-facing summaries unless they ask how the workflow is wired up.
≈ “Common” is an estimate: at least 1 in 5 of the 6,668 Cursor rule files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
--- description: Before substantial feature work, run ForgeTrail lessons + anti-patterns (MCP) and fold findings into the plan alwaysApply: true --- # ForgeTrail lessons gate (when + how) This complements **`forgetrail-lessons-mcp.mdc`** (tool details + code-path globs) and **`forgetrail-phase-status.mdc`** (phase footer). **Automation:** the IDE cannot invoke MCP tools without the agent doing it—this rule makes that **routine** for the right tasks. ## When to run the gate (do it early, before large edits) Run **`getAntiPatterns`** once and **`searchLessons`** (one or more queries) when **any** of these is true: - You expect to touch **≥3 files**, or **≥2** top-level areas (e.g. UI routes + server API + persistence). - **New or changed** user-facing flow, export/document pipeline, LLM prompt or response schema, or API route behavior. - The user asks for a **plan**, **feature**, **refactor**, or **milestone** that clearly spans multiple files. - **Schema / persistence** changes (validation shapes, database fields, migration scripts). **Skip the gate** for: single-file typos; renames with no behavior change; dependency version bumps with no UX/API impact; or when the user explicitly asks to skip or needs an emergency one-file hotfix. ## How to run it (order) 1. If the client exposes tool schemas, use them for **`searchLessons`** (query + optional limit) and **`getAntiPatterns`** (no arguments). In Cursor, descriptors may appear under the project’s `mcps/` folder for the configured ForgeTrail server. 2. Call **`getAntiPatterns`** (no args). Skim for items that apply to this task (structured docs vs LLM, vague errors, testing after refactor, schema drift, vague UI specs). 3. Call **`searchLessons`** with **2–4 short keyword queries** derived from the work (e.g. `wizard progress staged`, `Svelte $effect untrack`, `PPTX export`, `LLM JSON validate`). Use **keywords**, not full sentences; optional `limit` 8–12 if results are thin. 4. **Fold in before coding:** Add 3–6 bullet points to your plan or first implementation message—what you will **do** or **avoid** (e.g. staged progress for >3s flows, code-owned document structure, plain-language errors). Do **not** paste full MCP dumps unless the user asks for raw excerpts. ## Product-facing users Do not mention MCP, tool names, or “lessons gate” in UI copy or user-facing summaries unless they ask how the workflow is wired up.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.