agentleFS
lithammerlithammer/ai/skills/research-codebase/SKILL.md

research-codebase

Skill0 starsChanged 52 days agoMITView on GitHub

Map how the code a ticket touches works today, then name the next step: develop, refactor, or grill-with-docs.

9 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use validate-plan, iterate-plan and 4 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • shawn-dumasSkill

    validate-plan

    Validate an orchestration plan after pre-flight certification. Runs adversarial review, deep review (verify prompt data against codebase), and PoC gate for risky approaches. Auto-invoked by orchestrate-* skills at Step 8.

    Also in 4 projects that use this

  • corticalstackSkill

    iterate-plan

    Revise an existing implementation plan in flow/plans/ based on feedback, researching when needed. Explicit-only; run only when invoked via /iterate-plan or explicitly asked.

    Also in 3 projects that use this

  • parcadeiSkill

    describe-pr

    Generate comprehensive PR descriptions following repository templates

    Also in 4 projects that use this

  • JasonPaffSkill

    implement-plan

    Execute implementation plan with structured tracking and validation using subagent architecture

    Also in 4 projects that use this

  • JetBrainsSkill

    create-plan

    Research the codebase and create an implementation plan with architecture notes, design document, and track decomposition. Use when starting a new feature or large change.

    Also in 3 projects that use this

  • SentrySkill

    commit

    Create commit messages following Sentry conventions. Use when committing code changes, writing commit messages, or formatting git history. Follows conventional commits with Sentry-specific issue references.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 4 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 4

    Distinctive

    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
  1. Distinctive:

    Read the ticket and the files it names before dispatching anything.

    line 15
  2. Very specific:

    When the ticket names no area of code to change, there is nothing to map yet: route to grill-with-docs and stop.

    line 22
  3. Distinctive:

    Write one Markdown file:

    line 36
  4. Distinctive:

    Write every finding as an observation of the current code.

    line 57
  5. Distinctive:

    Before naming the route, open the cited lines for each fact it rests on.

    line 67
  6. Very specific:

    refactor: when the map traces the ticket's need to a cause outside the code the ticket names, and a structural change no larger than the …

    line 71
  7. Very specific:

    develop: when the ticket and the map leave one reasonable way to make the change: no open choice alters observable behavior or an …

    line 74
  8. Very specific:

    grill-with-docs: when a choice remains that is the user's to make.

    line 76

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “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

Research codebase

Map the code a ticket touches as it stands, so the next step starts from facts instead of finding them again. The map describes; the next step decides, so every choice stays open for the user.

1. Frame the questions

Read the ticket and the files it names before dispatching anything. Turn the ticket into questions about the current code: where the relevant parts live, how data and control flow through them, which callers, tests, and configuration depend on them, which patterns nearby code follows, and where the need the ticket addresses originates, such as a bug's cause. Each question asks how the code works now.

When the ticket names no area of code to change, there is nothing to map yet: route to grill-with-docs and stop.

2. Investigate

Dispatch a sub-agent for each independent question, so the main context holds the questions and findings, not the reading. Each sub-agent returns observations with path:line.

The step is done when every question has an answer backed by path:line, or is marked open with the places searched.

3. Write the map

Write one Markdown file:

# <topic>

Ticket: <reference>
Commit: <short sha>

## Questions

<numbered, as framed in step 1>

## Findings

<per question: what exists, where, how it works, how parts connect>

## Open

<questions the code did not answer, with the places searched>

Write every finding as an observation of the current code. A gap or quirk is a fact with a location like any other.

The map describes one commit and goes stale as the change lands. Save it where the repository keeps working notes for tasks; when it has none, under .research/ at the repository root, with .research/ listed in .git/info/exclude to keep it out of commits.

4. Route

Before naming the route, open the cited lines for each fact it rests on. Reply with a short summary of the map, its path, and the first route that applies:

  • refactor when the map traces the ticket's need to a cause outside the code the ticket names, and a structural change no larger than the ticket's own could remove the need.
  • develop when the ticket and the map leave one reasonable way to make the change: no open choice alters observable behavior or an interface.
  • grill-with-docs when a choice remains that is the user's to make. List each choice in one line.

The user starts the next step, often in a fresh session given the map's path.

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.