AGENT SKILL

setup-pstack

Configure which models pstack uses per role and at what reasoning budget. Detects your available models and writes a project models file that overrides the defaults. Use for /setup-pstack or $setup-pstack, "configure pstack models", "pstack budget", or changing pstack's model choices.

License 许可
MIT
最近更新
2026-10-28
Stars
★ 1
Forks
0

01 前言 Front Matter

name
setup-pstack
description
Configure which models pstack uses per role and at what reasoning budget. Detects your available models and writes a project models file that overrides the defaults. Use for /setup-pstack or $setup-pstack, "configure pstack models", "pstack budget", or changing pstack's model choices.

03 SKILL.md 正文

Setup pstack

Write .agents/pstack-models.md in the project's main checkout ("$(dirname "$(git rev-parse --path-format=absolute --git-common-dir)")/.agents/pstack-models.md", which may not exist). It sets pstack's model per role and overrides the defaults.

Steps

1. Detect available models

Call orchestrator_capabilities. Each entry in providers lists a provider instance's models and each model's effort options (effort, or reasoningEffort for codex). That is the dependable source. A provider whose canRunChildTask is false can't run a role, so leave it out. If you cannot detect any, ask the user to paste the models they have access to. Never write a real model you have not confirmed is available. The alias inherit-parent is always valid even though it is not a detected model.

2. Load current state

The defaults are the role entries listed in step 5 below. If that file already exists, read it and treat its # budget: line and its role entries as the current choices. Otherwise start from those defaults. A line whose role is not in step 5, such as how critics, is from a retired role. Drop it.

3. Budget, map, and confirm

(a) Ask for a budget. Ask in plain text, then end the turn. Offer these four options with these exact labels, and name the current budget when the file records one. With no models file, name the budget whose effort matches the defaults in step 5.

  • unlimited — max reasoning
  • large — xhigh reasoning
  • medium — high reasoning
  • small — medium reasoning

(b) Apply it. Build the working table from the defaults, and on a re-run keep any role you changed by model, list, or alias (inherit-parent). unlimited, large, medium, and small set the effort of every real entry, panel entries included, to max, xhigh, high, or medium, on the ladder max > xhigh > high > medium > low. Ignore effort options off that ladder. If the model doesn't offer that effort, use the highest effort it offers below it, else mark the role as needing a choice. inherit-parent does not change. So under unlimited, a model whose efforts stop at xhigh runs at xhigh.

(c) Show the roles and confirm. Show every role with its entries, marking any real entry not in the detected set as needing a choice. Also list each line step 2 dropped. Ask whether to accept as-is or change specific roles, offering the detected models plus inherit-parent (this role runs on the parent chat model) as the options. Ask in plain text, then end the turn. For panel roles (arena runners, architect runners, interrogate reviewers) the value is a list, and one subagent runs per entry, alias entries included, so the list length sets the count. arena cross-judge pool is also a list, but Arena selects one value from it whose model family differs from the parent's when possible. swarm workers is the default model for every worker unless a race or comparison assigns another model per arm.

4. Validate

Every real entry written must be a detected model with an effort that model offers. inherit-parent always passes. If a chosen real entry is not available, stop and ask again.

5. Write the models file

Write that file with a # budget: line holding the chosen label and its target effort, and one line per role, using the role names below. Create .agents/ if it is missing. Overwrite the whole file so re-runs stay idempotent. Shape:

# pstack model configuration. One line per role. Delete a line to fall back to the default.
# Entry: <provider>/<model> <effort>. `inherit-parent` runs the role on the parent's model (omit `target`). Alias entries in a panel list still count toward its fan-out.
# budget: <label> (<effort>)
<role>: <entry>, <entry>, ...

The roles, in this order, with their defaults:

  • feature, refactoring: claudeAgent/claude-sonnet-5-5 high
  • bug-fix: claudeAgent/claude-sonnet-5-5 high
  • perf-issue: claudeAgent/claude-haiku-5-5 xhigh
  • hillclimb: claudeAgent/claude-sonnet-5-5 high
  • judgment and prose: claudeAgent/claude-opus-5-5 high
  • hardest tasks: claudeAgent/claude-opus-5-5 xhigh
  • how explorer: claudeAgent/claude-haiku-5-5 xhigh
  • how explainer: claudeAgent/claude-opus-5-5 high
  • why investigators: claudeAgent/claude-haiku-5-5 xhigh
  • why synthesizer: claudeAgent/claude-opus-5-5 high
  • reflect tooling: claudeAgent/claude-haiku-5-5 xhigh
  • reflect judgment, divergent, synthesizer: claudeAgent/claude-opus-5-5 high
  • arena runners: claudeAgent/claude-opus-5-5 high, codex/gpt-6.1-sol xhigh
  • arena cross-judge pool: claudeAgent/claude-opus-5-5 high
  • swarm workers: claudeAgent/claude-haiku-5-5 xhigh
  • architect runners: claudeAgent/claude-opus-5-5 high, codex/gpt-6.1-sol xhigh
  • interrogate reviewers: claudeAgent/claude-opus-5-5 high, codex/gpt-6.1-sol xhigh

6. Confirm

Tell the user the file was written and that skills read it at each spawn. Re-running this skill updates it. Tell the user to commit the file if worktrees and other people should see it.

7. Offer a verification skill (optional)

Check whether the project has a way to drive the real app for proof (a verify-* skill, or an existing harness). If not, offer once: "want a project-local verification skill, so agents can drive the app the way a user does and prove changes work? I can generate one with the create-verification-skill skill." On yes, run the create-verification-skill skill (resolves wherever pstack is installed: workspace, user, or plugin). On no, move on without pushing.

04 同仓库更多 Skills

willgosnold
willgosnold/patchbook

patchbook-update

Pull upstreams into a repo that keeps its changes in a patchbook (.patchbook/), repair what broke from each change's intent, test and review, then merge and deploy only when the sync was clean or the test command passes, and otherwise ask, opening a pull request when there are commits to push. Use when a scheduled task or the user says to update a patchbook repo, or for /patchbook-update or $patchbook-update.

★ 1
willgosnold
willgosnold/patchbook

why

Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed thresholds. Discovers available MCPs and queries each evidence category (source control, issue tracker, long-form docs, real-time chat, infrastructure observability, error tracking, product analytics warehouse) in parallel, then returns a cited read on decisions and tradeoffs. Use how for runtime behavior.

★ 1
willgosnold
willgosnold/patchbook

verify-patchbook

Drive the patchbook CLI from this checkout in a disposable sandbox (its own git repos, store, config and a fake model) and capture a transcript as proof. Use to reproduce a patchbook bug, run an issue's repro script, or prove a change to clone, init, record, the pre-commit hook, build, status, check, sync or upstream works the way a user would see it.

★ 1
willgosnold
willgosnold/patchbook

unslop

Cut AI tells from any writing.

★ 1
willgosnold
willgosnold/patchbook

typescript-best-practices

TypeScript best practices. Use when reading or editing any .ts or .tsx file.

★ 1
willgosnold
willgosnold/patchbook

technical-writing

Layered technical-writing standard: Diátaxis structure, Google developer style sentences, STE instruction rules, Global English syntax. Use for /technical-writing or $technical-writing, or when writing or reviewing docs, RFCs, readmes, PR descriptions, or commit messages.

★ 1