agentleFS
cheyuriycheyuriy/codelabgen/skills/source-digest/SKILL.md

source-digest

Skill1 starsChanged 23 days agoMITView on GitHub

How to write the digest of one source for a codelabgen course (.codelabgen/sources/<id>.md) — a short, structured summary that later agents use to know what the source covers, where its examples are, and where to dig deeper when the course is extended.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 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.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write it in English, whatever the course language is.

    line 11
  2. Distinctive:

    Never pad. If a source turns out to be thin, say so in the summary.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Writing a source digest

A digest is a source's index card. Every later agent reads digests instead of the source itself: the curriculum architect to plan, the lesson writers to find material, and the gap analyst to extend the course months later. Make it accurate, specific and short: 30–80 lines.

Write it in English, whatever the course language is.

Format

.codelabgen/sources/<id>.md, with the id exactly as in codelabgen sources list:

---
id: official-docs--docs-langchain-com--3cfc8e
url: https://docs.langchain.com/oss/python/langgraph/graph-api
type: official-docs
roles: [code-truth, concepts]
title: Graph API overview
version: "1.x"
fetched: 2026-10-16
score: 0.95
---

## Summary
What the source is, its angle and its depth, in 2–4 sentences. Whom it is for, and how
current it is relative to the pinned version.

## Topics covered
- **StateGraph and state schemas**: TypedDict, Pydantic and dataclass schemas, input/output schemas ([#state](https://…#state))
- **Reducers**: `Annotated` with `operator.add`, `add_messages`, custom reducers (#reducers)
- …

## Code examples
- Minimal graph with two nodes and a conditional edge (#conditional-edges): complete, runnable
- `examples/persistence/postgres.py` in the repo: PostgresSaver setup with a connection pool
- …

## Gaps / dig deeper
- Skims the `Send` API; the full treatment is on the "Map-reduce" page (not fetched): https://…
- No coverage of async execution
- Follow-up queries: "langgraph durability modes", "langgraph subgraph state mapping"

Rules

  • Topics covered keeps the source's own order. For a book TOC or a course outline, that order is the point. Link sections with anchors or paths whenever you can, so a writer can jump straight there.
  • Code examples lists only examples that exist, with where to find them, and says whether they are complete and runnable.
  • Gaps / dig deeper is for the future: what the source skips or only mentions, linked pages you didn't read, versions it doesn't cover, and search queries that would continue the research.
  • Never pad. If a source turns out to be thin, say so in the summary. The orchestrator may drop it.
  • Copy the front matter fields from the registry entry. Set fetched to today.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.