agentleFS
orbyn-devorbyn-dev/orbyn-fullstack-skills/skills/why/SKILL.md

why

Skill4 starsChanged 10 days agoMITView on GitHub

Investigate design rationale, regressions and trade-offs using Git history and relevant authorized records; distinguish documented intent from inference.

112 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use addskill, apply-firmware 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.
  • ryanportfolioSkill

    addskill

    Create, import, update, or install repo or personal skills: runtime ownership, resources, discovery checks.

    Also in 6 projects that use this

  • ryanportfolioSkill

    apply-firmware

    Bring current folder (plain folder, existing repo, old or partial firmware copy) up to full Harness Firmware in place. Mirror external repo → adopt-repo.

    Also in 6 projects that use this

  • +10 more from ryanportfolio

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 105 edited versions we read keep these.

  • 4

    Distinctive

    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:

    Follow the consuming project's instructions and requested scope.

    line 8
  2. Distinctive:

    Read linked ADRs, issues, PR discussions and design records: only to answer the current question.

    line 15
  3. Distinctive:

    Do not create tickets or issues, message authors or send other messages while investigating or merely to complete a historical explanation.

    line 23
  4. Distinctive:

    No connector installation, capture or cross-account search follows.

    line 37

“Common” is measured: kept by at least 6 in 10 of the 105 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

Why

Follow the consuming project's instructions and requested scope; use the portable defaults only where guidance is missing.

Method

  1. Anchor the question in the changed code, symbol or behavior. Treat the user's proposed explanation as a hypothesis.
  2. Locate introducing and corrective commits with scoped git log, git blame and diffs. Read the relevant diff and surrounding source, because line attribution alone does not establish intent.
  3. Read linked ADRs, issues, PR discussions and design records only to answer the current question. Expand to other available authorized records only to resolve a specific gap. Do not search all connected accounts or private chats by default.
  4. Record direct rationale, contrary evidence, plausible alternatives and inference separately. Code establishes mechanics, not necessarily motivation. A later comment can describe current policy without proving why the original implementation was chosen.
  5. Return the supported explanation with sources, inference and unknowns. If work will change the design, name the constraints to preserve and the decisions to revisit. Describe the searched boundary: an empty bounded search means no rationale was found in that scope, and does not prove no record ever existed.

Take a worker that uses a lease instead of a process-local lock. Inspect the introducing change and related recovery contract. Source can establish cross-process expiry and recovery behavior; only a linked rationale or other authorized record can establish why the original author selected it. A linked design note may show crash-recovery motivation. If that note is absent, report the mechanism as confirmed and the motivation as inferred, with the searched history boundary.

Investigation limits

Do not create tickets or issues, message authors or send other messages while investigating or merely to complete a historical explanation.

For substantial historical investigation, the lead delegates bounded evidence questions to named Orbyn Investigators under the workflow, then inspects their cited records and owns the explanation. Narrow questions and children assigned a slice proceed directly.

Recover the introducing decision

When last-touch blame misses the origin, use scoped rename-following history and symbol/string introduction search (git log --follow for a selected file and pickaxe -S/-G for the relevant symbol), then inspect the introducing diff. Treat a later move/sync commit as a possible carrier rather than the decision or defect origin. Read material discussion, parent/duplicate chains and the finalized outcome before treating an issue title or preview as rationale; a final reversal can supersede its proposal. Bound the investigation by the unresolved question rather than scanning all history.

Authorized operational records

  • Distinguish a bounded empty query from missing authentication, expired retention or unavailable history.
  • Discover the actual tool/schema, select a relevant time/release/environment window and poll the same asynchronous query identity instead of resubmitting costly scans.
  • Preserve denominator, sampling, refresh and instrumentation/schema changes before interpreting volume as adoption or causality.
  • Representative primary events support relevance; generated diagnoses remain hypotheses.
  • No connector installation, capture or cross-account search follows.

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.