agentleFS
lithammerlithammer/ai/skills/refactor/SKILL.md

refactor

Skill0 starsChanged 52 days agoMITView on GitHub

Find structural refactors that remove caller obligations, then implement the chosen change.

196 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use ascii-diagram, debug-fix and 10 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.
  • SethGammonSkill

    ascii-diagram

    Generate perfectly aligned ASCII diagrams — architecture, flow, sequence, box-and-arrow. Uses a programmatic character-grid approach so alignment is guaranteed by math, not token prediction. Includes post-render verification.

    Also in 4 projects that use this

  • zebbernSkill

    debug-fix

    Find and fix a bug or issue — from any source (GitHub issue, error message, user report, or observed behavior)

    Also in 4 projects that use this

  • parcadeiSkill

    agent-context-isolation

    Agent Context Isolation

    Also in 3 projects that use this

  • parcadeiSkill

    agentica-infrastructure

    Reference guide for Agentica multi-agent infrastructure APIs

    Also in 3 projects that use this

  • +8 more from parcadei

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 169 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the repository's instructions and the requested scope or concern: a path, subsystem, goal such as making invalid checkout states …

    line 15
  2. Distinctive:

    Use a single agent by default.

    line 24
  3. Distinctive:

    Read each promising implementation and its callers in full, with the tests and contracts needed to judge the change.

    line 44
  4. Distinctive:

    Wait for the user to select a candidate and approve any changes to accepted behavior before editing.

    line 76
  5. Very specific:

    Use the develop skill for the chosen change.

    line 81

“Common” is measured: kept by at least 6 in 10 of the 169 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

Refactor

Explore existing code for structural changes that remove obligations from callers. No feature or diff is needed.

1. Scope

Read the repository's instructions and the requested scope or concern: a path, subsystem, goal such as making invalid checkout states unrepresentable, or a ticket and its map from research-codebase. With a ticket, inspect the code its change touches; a candidate's cost is what that change would otherwise have to work around. With no target, map the repository's major subsystems and choose promising areas to inspect.

2. Discover

Use a single agent by default. Apply these related questions in one pass; they overlap and are not separate work queues. Split a large scope only along genuinely independent subsystems.

  • Representation: which units, identities, states, or operations rely on primitives, flags, naming, or caller discipline?
  • Derivability: which stored or passed values could be derived from authoritative state instead of kept in sync?
  • Invariant ownership: which rules do callers repeatedly enforce that one owning module or storage constraint could guarantee?
  • Lifecycle ownership: which acquire/release or transaction protocols could a scoped operation own instead of each caller?
  • Least capability: which callers receive authority they do not need?
  • Altitude: which special cases compensate for limitations in the underlying mechanism? Where should the rule live?
  • Deep modules: which repeated orchestration could sit behind an interface that hides knowledge callers currently need?

3. Substantiate

Read each promising implementation and its callers in full, with the tests and contracts needed to judge the change. Trace the burden or possible misuse to concrete code. A candidate needs a present cost or credible failure mode; an observed bug is not required.

Describe the replacement and check which guarantees it actually enforces, and which still need validation or tests. Distinguish tighter internal representations from changes to accepted external behavior. Preserve required boundary validation.

Merge candidates that address the same underlying mechanism.

4. Present and wait

Rank candidates by concrete benefit and migration risk. With a ticket, rank first a candidate that removes the ticket's need at its cause, making the ticket's change unnecessary. For each, give:

  • Locations and evidence of the current burden or possible misuse.
  • The proposed representation or owning module, and what callers stop doing.
  • Expected size (small/medium/large), justified by the affected modules, callers, and migration work. State any uncertainty.
  • Guarantees gained, affected contracts, migration risks, and verification needed.

State what you explored and what remains unread. Return no candidates when none meet the evidence bar.

Render the candidates as an HTML report following HTML-REPORT.md, and reply with its path and the ranked candidate titles, one line each.

Wait for the user to select a candidate and approve any changes to accepted behavior before editing.

5. Implement

Use the develop skill for the chosen change. Pass the selected candidate's evidence, scope, accepted contracts, and risks as its requirements. With a ticket, finish the refactor before the ticket's change, so each lands as its own diff. When the refactor removes the ticket's need, verify the need is met and drop the ticket's change.

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.