agentleFS
Sequela02Sequela02/canasta/skills/workflow/config-projection-review/SKILL.md

config-projection-review

Skill0 starsChanged yesterdayMITView on GitHub

Review canonical configuration, generated or synchronized copies, ownership, drift, and safe convergence.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    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:

    Use when one configuration source generates or synchronizes other files, settings or installed copies.

    line 12
  2. Distinctive:

    Prefer the project's existing status evidence over a competing scanner.

    line 19
  3. Distinctive:

    If no shared status exists, inspect the relevant files directly and state the limits.

    line 20
  4. Distinctive:

    Do not mutate during review.

    line 28

≈ “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

Config Projection Review

Use when one configuration source generates or synchronizes other files, settings or installed copies. A projection is a derived copy, not a second source of truth. This applies to build configuration, editor settings and agent instructions alike; it does not require a particular configuration service.

  1. Inventory the canonical sources, generator or synchronization mechanism, revisions, source digests, current target bytes and ownership evidence. Prefer the project's existing status evidence over a competing scanner. If no shared status exists, inspect the relevant files directly and state the limits; do not infer ownership merely from a matching filename.
  2. Classify each source and target as canonical, current, missing, stale, drifted or unmanaged. Record which fields or files are managed, source and target identities, comparison evidence and unreadable or unresolved state.
  3. Separate incorrect source content from incorrect synchronization. Use documentation-hygiene to assess repository guidance and its source facts; this review owns whether the selected derived copies can converge safely.
  4. Do not mutate during review. Preserve unrelated fields and unmanaged content. Block automatic repair when ownership is ambiguous, source configuration is invalid, inventory is incomplete, permission has changed, or a safe merge has not been established.
  5. Recommend the narrow owning action: correct the source if needed, preview the target diff, then synchronize only explicitly selected targets under the project's authorization rules. Recheck the source and target before applying an approved change; stale approval must not overwrite newer work. Do not recommend hand-editing a generated copy to hide a canonical-source defect.
  6. Account for partial failure per target. Say whether convergence is safe, needs review or is blocked. One unreadable or unsettled target prevents a claim that the whole operation succeeded.

Report findings first, overall status (valid, needs_changes or blocked), per-target evidence and owning action, verification performed and residual unknowns. A proposed repair is not permission to execute it.

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.