integrate-learnings

A command for processing community learning contributions into reference documentation. The contributions are JSONL files, meaning files containing one JSON object per line, and the command places eligible lessons into the appropriate guides.

In plain words
What is it for?
Use it in the mathlib-quality repository to preview or integrate contributions into style rules, naming conventions, proof patterns, and pull-request feedback examples.
Why use it?
It turns scattered contributor feedback into shared documentation that future workers can use as the source of truth.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/cbirkbeck/mathlib-quality/integrate-learnings
Clone the repo
git clone --depth 1 https://github.com/CBirkbeck/mathlib-quality
Per session 12 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,892 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00012 $0.01892
Opus 5 $0.00006 $0.00946
Sonnet 5 $0.00002 $0.00378
Haiku 4.5 $0.00001 $0.00189

Measured 2d ago against content hash 7ea3a693d0cd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

integrate-learnings scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/integrate-learnings.md · 212 lines

How it starts

The opening of the file, as written. The whole thing — 212 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/integrate-learnings — Integrate community learnings into the reference docs

For repo maintainers: process community learning contributions (from /contribute PRs) into the skill's reference documentation. Contributions land as JSONL files in data/community_learnings/; this command reads them, counts occurrences, and propagates the eligible ones into references/style-rules.md, references/naming-conventions.md, references/proof-patterns.md, references/pr-feedback-examples.md, etc. — wherever the teaching belongs.

Workers read the reference docs directly. There is no RAG index, no MCP server, no merge_learnings.py; the reference docs are the single source of truth.

Usage

/integrate-learnings
/integrate-learnings --dry-run   (preview without modifying files)

Prerequisites

  • Must be run from within the mathlib-quality repo itself
  • Community learnings exist in data/community_learnings/*.jsonl

Workflow

Step 1 — Load all contributions

ls data/community_learnings/*.jsonl

Parse each file (JSONL: one JSON object per line). Skip .gitkeep.

If no files found:

No community learnings to process in data/community_learnings/.
Contributions arrive via `/contribute` PRs from users.

Step 2 — Validate and deduplicate

For each entry:

  1. Validate JSON structure — must have required fields (command, type, description).
  2. Dedup — same before_code + after_code is a duplicate (keep the most recent). For entries without code (pure teachings), dedup by description prefix.
  3. Filter — drop entries with empty descriptions or missing type.

Report:

## Validation Report
- Total entries: 150
- Valid: 142
- Invalid (missing fields): 5
- Duplicates removed: 3
- Ready to integrate: 134

Step 3 — Count occurrences (consensus filter)

Group validated entries by pattern_tags. Print the occurrence count:

| Pattern (representative description)  | Occurrences | Sources                                   | Eligible for ref docs? (≥3) |
|---------------------------------------|-------------|-------------------------------------------|------------------------------|
| junk_def / inline_def                 | 4           | 2026-05-03_mathlib4, 2026-04-21_mathlib4  | yes                          |
| splits_api / unary_predicate          | 3           | 2026-05-06 mathlib-bump session (×3)      | yes                          |
| review-meta-pattern                   | 1           | 2026-04-29 community contribution         | no — single source           |

Read the full file on GitHub · 212 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 212 lines · 12 tokens per session scan A 7ea3a693d0cd

Subscribe to this mod's changes

integrate-learnings is a command published in the GitHub repository CBirkbeck/mathlib-quality (32 stars, last pushed 13d ago), licensed MIT. It adds 12 tokens to every session and 1,892 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.