agentleFS
jisuzzzzjisuzzzz/haechi/skills/spec/SKILL.md

spec

Skill17 starsChanged 14 days agoMITView on GitHub

Establish the problem, scope, expected behavior and acceptance criteria, with an optional saved spec. Use when the user requests a spec. Strongly recommended for new or changed workflows and interacting behavior such as permissions, validation, failure handling or cross-system contracts, to bring scope, edge cases and checkable completion criteria together before implementation, even when individual requirements are known. Also recommended in other playbooks when expected behavior, preserved contracts, scope or success criteria need clarification. Requirements can be known but still need synthesis. A narrow task with sufficiently established requirements and completion criteria can proceed directly without this skill; it does not require a document, repeated questions or approval.

168 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use backprop, ios-clean 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.
  • JuliusBrusseeSkill

    backprop

    Bug → spec protocol. When a bug is found or a test fails, trace the cause, decide whether a new §V invariant would catch recurrence, append to §B. This is the one non-obvious thing SDD does that plan-then-execute doesn't. Triggers on test failure, bug report, post-mortem, or explicit user ask.

    Also in 4 projects that use this

  • Saqlain-811Skill

    ios-clean

    Remove the DebugBridge SPM package and all #if DEBUG wiring from an iOS app. (gstack)

    Also in 5 projects that use this

  • Saqlain-811Skill

    ios-fix

    Autonomous iOS bug fixer. (gstack)

    Also in 5 projects that use this

  • bastosSkill

    ios-design-review

    Review iOS app UI/UX design quality using Apple's design principles and the Liquid Glass-era design language from WWDC25/26. Use when asked for a design review, UX critique, visual design assessment, or evaluation of an app's interface quality, design intent, alignment with the new Apple design system, or usability. Triggers: "design review", "UX review", "visual design critique", "review the UI", "does this design work", "evaluate this interface", "design feedback", "Liquid Glass review".

    Also in 5 projects that use this

  • ShreyRaviSkill

    plan-tune

    Self-tuning question sensitivity + developer psychographic for gstack (v1: observational). Review which AskUserQuestion prompts fire across gstack skills, set per-question preferences (never-ask / always-ask / ask-only-for-one-way), inspect the dual-track profile (what you declared vs what your behavior suggests), and enable/disable question tuning. Conversational interface — no CLI syntax required.

    Also in 7 projects that use this

  • timurgaleevSkill

    document-generate

    Generate missing docs for a feature, module or project as tutorials, how-tos, reference and explanation.

    Also in 6 projects that use this

  • +2 more from Saqlain-811

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 145 edited versions we read keep these.

  • 5

    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. Distinctive:

    Only the main session runs this skill.

    line 10
  2. Distinctive:

    Resolve consequential decisions: Read the code, docs and prior conversation first.

    line 19
  3. Distinctive:

    Optional saved spec: Write Markdown directly using the repository's document convention.

    line 21
  4. Very specific:

    Use Draft for proposed decisions and Approved only when the conversation establishes the user's acceptance.

    line 23
  5. Distinctive:

    Proceed on existing authorization: Ask only about consequential unresolved decisions or actions outside the user's authorization.

    line 25
  6. Very specific:

    When saving a spec, honor an explicit path, the repository convention or specDir from .haechi.json.

    line 30
  7. Distinctive:

    After implementation, update the feature's canonical doc with the changed behavior and decisions, preserving unchanged contracts and …

    line 31

“Common” is measured: kept by at least 6 in 10 of the 145 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

Write a spec

Apply the common work contract within this task’s scope. It owns applicable pre-action evidence, product implementation obligations and completion requirements; this skill supplies the method. Current common guidance already delivered or read may be reused.

Only the main session runs this skill. Subagents cannot ask the user anything, so they never write a spec; they report open questions to you.

A spec answers what and why: problem, goal, scope, expected behavior, edge-case decisions, acceptance criteria, constraints. Requirements already established in the conversation or repository can serve as the spec; they need no repeated approval. A plan answers how and in what order: files, type sketches, units, delegation, checks, risks. Choose the implementation plan yourself.

Selection. Use the conditions in this skill's description. This is a recommendation, not a mandatory stage in every feature or other playbook. Do not equate known individual requirements with a sufficiently established overall scope and acceptance criteria. Conversely, reuse a complete existing account of the requirements and proceed directly when further synthesis adds no value. File counts, code size, a new screen or a changed export alone do not decide selection. Selecting this skill does not require a saved document, repeated questions or approval. Resolve consequential open decisions before dependent work even when this skill is not selected; decide reversible implementation details yourself.

Steps

  1. Establish the requirements. Identify the intended outcome, scope, behavior that must stay unchanged and acceptance criteria. A brief conversation summary is enough unless a saved spec is requested or useful. Fixed labels and tables are optional.
  2. Resolve consequential decisions. Read the code, docs and prior conversation first. Batch the questions that remain and explain meaningful alternatives. Reuse accepted decisions without asking again. Continue work that does not depend on an unanswered question.
  3. Visual previews. For new or changed visible UI or user interactions, run /haechi:preview before implementation, even without an explicit preview request. Also use it for requested mockups or visual variants that clarify a consequential choice. Reuse the user's existing preference; a preview does not require a saved spec.
  4. Optional saved spec. Write Markdown directly using the repository's document convention. Include the problem, goal, scope, expected behavior and acceptance criteria; add edge cases, constraints, open questions and decisions when relevant. Keep one editable document and omit empty sections.
    • Each acceptance criterion must be checkable: a test name, a screen state or an observable response.
    • Use Draft for proposed decisions and Approved only when the conversation establishes the user's acceptance. Do not require another approval for an accurate record of already accepted requirements.
  5. Review the wording. For a saved draft, apply the common writing rules before showing it; consult the clear-writing principle as needed. Apply edits directly to the Markdown document.
  6. Proceed on existing authorization. Ask only about consequential unresolved decisions or actions outside the user's authorization. Do not add a formal approval round merely because a document was written. When the user explicitly requests a review before implementation, wait for that review.
  7. Plan. Choose the implementation order and delegation. Summarize it when useful; it needs no separate approval.

Where it lives

  • When saving a spec, honor an explicit path, the repository convention or specDir from .haechi.json; fill placeholders such as {feature} from the task. Missing specDir does not block work or require setup. Choose a reversible location inside the repository when needed.
  • After implementation, update the feature's canonical doc with the changed behavior and decisions, preserving unchanged contracts and exception cases. A temporary dated spec may be removed after merge when cleanup is authorized; the history stays in the commits.

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.