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

request

Skill17 starsChanged 14 days agoMITView on GitHub

Use to write or revise a change request, fact question or bug report for another team, or to incorporate answers into an existing request. Keeps replies, decisions and unresolved questions in the same document. Honors the user's format and destination; defaults to their notes folder. Writing does not authorize sending.

18 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 17 edited versions we read keep these.

  • 1

    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:

    Keep requests outside repositories by default.

    line 27

“Common” is measured: kept by at least 6 in 10 of the 17 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

Request to another team

Example: the order list needs a currency field that GET /api/orders does not return. The feature playbook stops and asks instead of working around it in the client. The user says "let's ask the API team for it". This skill writes that request.

Write one document that the other team can answer without this chat. Use the document skill for the shared Parts, situational templates, narrative, evidence and scoped human review. Keep this skill's notes location and same-file update behavior below. Default to Markdown; honor an explicitly requested format.

Draft

  1. Check what exists first. Read the other side's API schema, docs or code when you can reach them. Establish what is missing, which fact needs confirmation or which behavior was observed, as appropriate to the request. If that side is inaccessible, state what remains unconfirmed. Put detailed evidence in endnotes, with a reference beside the supported claim.
  2. Choose and fill the shared template in <notesDir>/requests/YYYY-MM-DD-<slug>.md. A new API or field normally uses change; an unanswered fact uses fact; an observed error uses bug. Apply the conditional Parts that the case needs. Start with the current need; no old document is required. Use summary, linked TOC, numbered sections and the human review record from the document skill. Use tracking IDs or a table only when they help this particular request.
  3. Make each item answerable without a meeting. State what is needed, why it matters, its priority and current status, then give:
    • the expected contract: response shape, field meanings, nullability and an example payload;
    • proof of completion, for example "GET /api/orders/42 returns "currency": "EUR"";
    • alternatives only when the request is costly, and deployment order only for a breaking change. Include a due date only when one is established. Track unanswered questions and dated decisions when they exist; omit empty sections. Suggested item statuses are Pending, Accepted, Accepted with changes, Rejected and On hold; translate them for the team when helpful.
  4. Collect visual evidence when needed using the shared capture procedure. Reuse browser helpers or temporary Playwright, preserve the original PNG, make a separate numbered-border PNG and attach both with explanations. GIF is optional when motion matters. Complete the document skill's plain-writing, navigation, evidence and image checks. Human review stays pending until the actual reviewer confirms its scope.

Where it lives

  • Save it as <notesDir>/requests/YYYY-MM-DD-<slug>.md, with notesDir from ~/.haechi/user.json. If notesDir is missing, ask once where notes go.
  • An explicit user-specified format or destination takes precedence over these defaults.
  • Keep requests outside repositories by default. They often contain chat quotes, parts of the other side's code and discussion of production systems.
  • The user asking for the request covers writing this file. Sending it — by chat, email, issue or anywhere else — requires authorization for that action. Reuse explicit authorization already given; otherwise show the file and ask before sending.

When answers come back

The user pastes or relays an answer, from chat or anywhere else. Update the same Markdown document directly:

  • the answer of each matching open question,
  • a dated decision for each resolved item,
  • each item's status in its summary and details,
  • the overall status: In progress once the other team starts answering, Done when no item is Pending or On hold.

Done here describes the request conversation. Keep implementation, deployment, verification and human review statuses separate. Use the document skill to refresh the summary and next action, preserve actual review scope/version, and record only real replies and decisions.

Then tell the user what changed for the feature: what is now unblocked, what still waits, and any change to the plan.

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.