agentleFS
PromptPartnerPromptPartner/agentsmith/skills/grill-with-docs/SKILL.md

grill-with-docs

Skill173 starsChanged 17 days agoMITView on GitHub

Grill a fuzzy, single-session repository change into shared language, recorded decisions, and a handoff-ready plan. Use when the operator asks to grill or stress-test a plan or design in a repository; route multi-session uncertainty to wayfinder.

195 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use setup-matt-pocock-skills, decision-mapping 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.
  • GoogleSkill

    setup-matt-pocock-skills

    Configure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.

    Also in 185 projects that use this

  • FabienDehopreSkill

    decision-mapping

    Turn a loose idea into a sequenced map of investigation tickets, then drive them to resolution one at a time.

    Also in 9 projects that use this

  • LesaniSkill

    writing-beats

    Shape an article as a journey of beats, choose-your-own-adventure style. The user picks a starting beat from the raw material, you write only that beat, then offer options for where to pivot next, beat by beat, until the article reaches a natural end. Use when the user has raw material and wants to assemble it as a narrative rather than an argument.

    Also in 65 projects that use this

  • liufeng1976Skill

    setup-ts-deep-modules

    Wire dependency-cruiser into a TypeScript repo so each package is a deep module — implementation hidden in subfolders, reachable only through its entry-point files. User-invoked.

    Also in 42 projects that use this

  • simo97Skill

    migrate-to-shoehorn

    Migrate test files from `as` type assertions to @total-typescript/shoehorn. Use when user mentions shoehorn, wants to replace `as` in tests, or needs partial test data.

    Also in 67 projects that use this

  • LesaniSkill

    writing-shape

    Take a markdown file of raw material and shape it into an article through a conversational session — drafting candidate openings, growing the piece paragraph by paragraph, arguing about format (lists, tables, callouts, quotes) at each step. Use when the user has a pile of notes, fragments, or a rough draft and wants help turning it into something publishable.

    Also in 63 projects that use this

  • +1 more from Lesani

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 181 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this workflow when a repository change is fuzzy but can be settled in one session.

    line 14
  2. Very specific:

    Use wayfinder when the destination spans sessions, has dependent decisions that need separate research, or cannot yet become one …

    line 15
  3. Distinctive:

    Use the ordinary plan flow when scope, language, and consequential choices are already settled.

    line 17
  4. Very specific:

    Read the tracked item, existing plan/spec, CONTEXT-MAP.md, relevant CONTEXT.md files, ADRs, and the code or assets that constrain the …

    line 21
  5. Distinctive:

    Ask a small numbered round from the frontier.

    line 31
  6. Distinctive:

    After each answer round:

    line 34
  7. Very specific:

    When a project-specific term resolves, update the relevant CONTEXT.md immediately using CONTEXT-FORMAT.md.

    line 54
  8. Very specific:

    Create an ADR only when the choice passes all three gates in ADR-FORMAT.md: hard to reverse, surprising without context, and the result of …

    line 60

“Common” is measured: kept by at least 6 in 10 of the 181 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

Grill with docs

Resolve a change's design tree with the operator while recording each answer where a fresh session can use it. Stop at a handoff-ready plan; implementation is a separate unit of work.

Route

  • Use this workflow when a repository change is fuzzy but can be settled in one session.
  • Use wayfinder when the destination spans sessions, has dependent decisions that need separate research, or cannot yet become one implementation plan.
  • Use the ordinary plan flow when scope, language, and consequential choices are already settled.

Establish the tree

  1. Read the tracked item, existing plan/spec, CONTEXT-MAP.md, relevant CONTEXT.md files, ADRs, and the code or assets that constrain the change. Facts come from the repository; decisions come from the operator.
  2. Locate the project's plan convention. Reuse an existing change plan; otherwise create one from .harness/templates/plan.md, templates/plan.md, or the same shape at .planning/<slug>.md, in that order of availability.
  3. Map the unresolved choices as a design tree. Its frontier is the independent questions whose prerequisites are settled now.

Work the frontier

Ask a small numbered round from the frontier. For each question, state why it matters, the real options and trade-offs, and your recommended answer. Then wait for the operator's decisions.

After each answer round:

  1. Recompute the tree. Cross-check claims against the repository and surface contradictions.
  2. Record every settled decision in the plan with its original precision: boundaries, ordering, defaults, failure behaviour, numeric constraints, and evidence requirements.
  3. Record resolved project language and qualifying ADRs before the next question round.

Repeat until the frontier is empty. A question that depends on an unresolved answer stays off the current frontier.

Durable artifacts

Plan

The plan is the complete decision ledger and implementation handoff. Keep its scope, approach, risks, verification, and done criteria current after each round. Place rationale once: in the plan for ordinary choices, or in an ADR for a qualifying architectural choice with a link from the plan.

Project language

When a project-specific term resolves, update the relevant CONTEXT.md immediately using CONTEXT-FORMAT.md. A root CONTEXT-MAP.md selects the bounded context; without one, use the root glossary. Create the file lazily on the first qualifying term.

Architectural decisions

Create an ADR only when the choice passes all three gates in ADR-FORMAT.md: hard to reverse, surprising without context, and the result of a real trade-off. Create docs/adr/ lazily and preserve the repository's existing ADR convention when one exists.

Completion

Offer the plan for confirmation only when the frontier is empty, every settled decision is durable, repository contradictions are resolved or explicit, and the plan defines end-to-end evidence. The session completes when the operator confirms shared understanding. Report the plan path, glossary and ADR changes, and whether the next route is implementation or wayfinder.


Adapted from Matt Pocock's grilling, domain-modeling, and grill-with-docs skills (MIT, 2026). AgentSmith makes the workflow self-contained and keeps every settled decision in the repository 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.