Agent skill

Squid Plan

by iusztinpaul in iusztinpaul/squid

Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides…

Apache-2.0Auto-check passedDevelopment

Install Squid Plan

skills CLI
$ npx skills add iusztinpaul/squid --skill squid-plan -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install iusztinpaul/squid squid-plan --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/iusztinpaul/squid.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/squid-plan .claude/skills/squid-plan && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
squid-plan
GitHub stars
203
Token cost
~2.6k tokens
SKILL.md length
1,150 words
Files
1
Skills in repo
13
Repo updated
First seen
Licence
Apache-2.0

At a glance

Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides…

  • Works in 6 steps: Resolve the feature spec → Grill the spec (Human ↔ /squid-grilling) → PA grooms → DRAFTS the plan (no writes… → …
  • Tasks that involve Architecture decision records
  • SKILL.md covers Step 0 — Resolve the feature…, Step 1 — Grill the spec (Human…, Step 2 — PA grooms → DRAFTS… and Step 3 — Offer another…, plus 3 more sections
  • Calls git and gh

What it does

Squid Plan is an agent skill from iusztinpaul/squid. Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides everything touching the repo: tasks + storage, ADR, glossary, worktree, and which build to run.

Its SKILL.md is about 2.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Development, covering Architecture decision records, PRD writing and Git worktrees. The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Architecture decision records
  • Tasks that involve PRD writing
  • Tasks that involve Git worktrees

Example prompts

  • “/squid-plan”

Workflow steps

6 steps, taken from the step headings in SKILL.md.

  1. Resolve the feature spec
  2. Grill the spec (Human ↔ /squid-grilling)
  3. PA grooms → DRAFTS the plan (no writes yet)
  4. Offer another grilling round, then output the final plan (Human ↔ /squid-grilling)
  5. HUMAN GATE (blocking)
  6. Execute the decisions (only after the final plan + Approve)

What it can do on your machine

Read from SKILL.md and the folder at commit f5bf6b3. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • gh

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git and gh, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Squid Plan loads about 2.6k tokens when it runs. Until then it costs about 75 tokens; SKILL.md has 1,150 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~75
When it runs · the whole SKILL.md, loaded when a task matches
~2.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from iusztinpaul/squid at commit f5bf6b3, republished under its Apache-2.0 licence (© iusztinpaul). 1,150 words, ~2,623 tokens.

Download SKILL.mdSave it as .claude/skills/squid-plan/SKILL.md (or your agent's skills folder).
name
squid-plan
description
Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides everything touching the repo: tasks + storage, ADR, glossary, worktree, and which build to run.
disable-model-invocation
true
argument-hint
<feature-spec | path/to/spec.md | tracker-ref>

Plan — feature spec → per-task files (+ optional ADR, + worktree)

Anchor a feature in shared understanding and prior decisions, then produce the artifact /squid-implement-night consumes. Nothing touches the repo until the human gate.

$ARGUMENTS is the raw feature spec — free-form text, a path to a spec file, or a tracker reference.

You are the orchestrator — a MANAGER. You drive the grilling, launch the Product Architect (PA), offer another grilling round and present the final plan, run the single human gate, set up the workspace, write the approved artifacts, and kick off the chosen build. You do NOT groom, write code, or implement anything yourself.

Read AGENTS.md first to confirm the active tracker mode (file → tasks/<NNN>-<slug>.md files; gh → one GitHub Issue per task).

Input: raw feature spec. Output: decided at the human gate — by default one tasks/<NNN>-<slug>.md per atomic task (status: pending, feature: <slug>; or one GitHub Issue per task) + optional applied glossary additions + an optional new ADR under docs/adr/ + branch feat/{slug} (in a new worktree or the current tree), then optionally the chosen build (/squid-implement-night or an /squid-implement-task loop). This is exactly what the downstream pipeline consumes.


Step 0 — Resolve the feature spec

If $ARGUMENTS is empty, ask the human for the feature (free-form, path, or tracker ref). Otherwise resolve it: cat a spec file, load a tracker record, or use free-form text. Surface the resolved spec back in one paragraph.


Step 1 — Grill the spec (Human ↔ /squid-grilling)

Before grooming, sharpen the raw spec with the human. Invoke the squid-grilling skill — interview the human relentlessly, one question at a time with a recommended answer, until scope, edge cases, non-goals, constraints, and any decisions that warrant an ADR are clear. Anchor the questions in what already exists: read docs/adr/ (settled decisions — don't re-open them) and docs/glossary.md (use its terms; flag any the spec uses differently) when present, and use the context7 plugin for authoritative library/API facts. Anything answerable by reading the codebase or those sources — explore instead of asking. The output is a grilled spec; carry it into Step 2.


Step 2 — PA grooms → DRAFTS the plan (no writes yet)

Launch ONE Product Architect. It drafts everything and hands it back as its final message — it writes nothing to disk; the orchestrator writes the approved artifacts into the chosen workspace in Step 5.

Agent(
  subagent_type="squid:product-architect",
  prompt="""Feature-level grooming. Read AGENTS.md first. Follow your feature-grooming role.
  Feature (grilled): {grilled spec from Step 1}.
  Decompose into atomic, independently-shippable tasks, numbered (NNN) in dependency order. For EACH task, draft the
  FULL tasks/<NNN>-<slug>.md content (`feature: {slug}`, `status: pending`) per the tracker-workflow spec.
  Also draft: (a) any new docs/glossary.md terms, and (b) IF the feature warrants non-obvious architectural decisions,
  ONE proposed ADR for the WHOLE feature per your ADR rule.
  DO NOT WRITE ANYTHING TO DISK — hand everything back as drafts; the human approves and the orchestrator writes them.
  Use the context7 plugin for authoritative library/API usage wherever the feature touches an external framework.
  Return: (1) the ordered task files with their full content, (2) the glossary additions (or 'none'), (3) the proposed
  ADR draft (or 'none')."""
)

Verify before the gate: the drafts are atomic, ordered by NNN in dependency order, and each has acceptance criteria. Re-launch the PA with the gap as feedback if not.


Step 3 — Offer another grilling round, then output the final plan (Human ↔ /squid-grilling)

Before /squid-plan moves toward implementation, give the human one explicit chance to sharpen further — then lock and show the final plan.

Ask once with AskUserQuestion: "Another grilling round to sharpen this, or is the plan final?" → More grilling / It's final.

  • More grilling → re-invoke the squid-grilling skill on the points the drafts left open, feed the sharpened spec back to the PA (Step 2), and return here. Repeat until the human picks It's final. Keep it to genuinely-open questions — good: "should deleting an Order cascade to its line-items?"; bad: re-opening a settled choice like "maybe switch datastores after all" (that's a fresh /squid-plan, not another round).
  • It's final → output the final plan in full so the human reads exactly what will be built before anything is written: every task (NNN — title, scope, acceptance criteria, out-of-scope), the glossary additions, and the proposed ADR. This is the human's last look before tasks are created — render it complete, not a teaser.

Why this step: catching a wrong-shaped plan on screen costs one more grilling round; catching it after tasks, a branch, and a build already exist costs the whole downstream pipeline.


Show full SKILL.md (619 more words)Show less

Step 4 — HUMAN GATE (blocking)

This is the only gate, and it is mandatory. It decides everything that touches the repo. Recap the decision inputs, then ask:

Feature: {title}
Tasks ({N}):
  1. {NNN-slug} — {title}
  2. ...
Glossary additions: {M new terms, or "none"}
Proposed ADR: {ADR-NNNN "title" — 1-line summary, or "none"}
Task storage (project default): {file → local tasks/*.md | gh → GitHub Issues}   ← from AGENTS.md TRACKER_MODE

Then ask with AskUserQuestion. The decisions come in two back-to-back asks — still ONE gate, with no repo writes between them: Part 2 only matters once the plan is approved, and AskUserQuestion caps at four questions per call.

Part 1 — the plan and its artifacts (ask together, then act on Q1):

  1. Approve the plan? — write these {N} tasks? → Approve / Edit / Cancel
  2. Store the tasks where? — → Local files (tasks/<NNN>-<slug>.md, committed to the repo) / GitHub Issues (one issue per task). Pre-select the project default (AGENTS.md TRACKER_MODE) and mark it Recommended.
  3. Create the ADR? — write the proposed ADR to docs/adr/? → Create / Skip (omit this question entirely if no ADR was proposed)
  4. Update the glossary? — apply the {M} drafted term(s) to docs/glossary.md? → Apply / Skip (omit this question entirely if the PA drafted no glossary additions)
  • Cancel → stop and discard the drafts; nothing has been written or branched. Do not ask Part 2.
  • Edit → ask what to add / remove / reorder / re-split; re-launch the PA (Step 2); loop back to this gate.
  • Approve → ask Part 2, then go to Step 5 carrying every answer.

Part 2 — workspace and build (only on Approve):

  1. Workspace? — where should branch feat/{slug} live? → New worktree (isolated ../{repo}-{slug} dir) / Current working tree (branch in place)
  2. Build now? — what runs after setup? → /squid-implement-night (end-to-end to a validated PR) / /squid-implement-task loop (build + commit the tasks only, no review/CI) / Stop after planning

Step 5 — Execute the decisions (only after the final plan + Approve)

Do these in order.

A. Set up the workspace (Q5).

  • New worktree:
    bash
    git rev-parse --abbrev-ref HEAD          # expect main; if not, ask the human how to proceed
    git pull
    WORKTREE_PATH="$(git rev-parse --show-toplevel)/../$(basename $(git rev-parse --show-toplevel))-{slug}"
    git worktree add -b feat/{slug} "$WORKTREE_PATH" main
    If it already exists (re-running after an abort): tell the human, ask reuse (r) / recreate (d) — default reuse.
  • Current working tree:
    bash
    git pull
    git switch -c feat/{slug}                # if already on a feat/* branch, reuse it instead
    WORKTREE_PATH = the repo root.

B. Write the approved artifacts into the workspace. Write to absolute paths under $WORKTREE_PATH (for a new worktree the orchestrator's cwd is still the main repo — do NOT write the tasks into main):

  • Tasks (Q2). Local files → one $WORKTREE_PATH/tasks/<NNN>-<slug>.md per task (status: pending) from the Step 2 drafts. GitHub Issues → gh issue create one issue per task from the same drafts, in NNN order (titles NNN — {slug}, body = the groomed spec). The chosen mode is this feature's tracker for the rest of the pipeline; if it differs from AGENTS.md TRACKER_MODE, it's a one-off override for this plan — don't rewrite AGENTS.md.
  • Glossary (Q4). If Apply: apply the drafted additions to $WORKTREE_PATH/docs/glossary.md. If Skip (or none were drafted): discard them.
  • ADR (Q3). If Create: write $WORKTREE_PATH/docs/adr/NNNN-<slug>.md (Status: Accepted) from the ADR draft. If Skip: discard the draft.
  • Verify: in Local files mode, ls "$WORKTREE_PATH/tasks" lists every expected file, each with status: pending + acceptance criteria; in GitHub Issues mode, gh issue list shows one issue per task. If anything is missing, write it now — do not hand off an empty plan.

C. Kick off the build (Q6).

  • /squid-implement-night → invoke /squid-implement-night with: feature {slug}, Working directory: $WORKTREE_PATH.
  • /squid-implement-task loop → invoke /squid-implement-task with: the feature's pending tasks (tasks/<NNN>-*.md, status: pending), Working directory: $WORKTREE_PATH.
  • Stop after planning → hand off and stop:
    Plan approved. {N} tasks in tasks/ (status: pending) on `feat/{slug}` ({worktree at $WORKTREE_PATH | current working tree}).
    Next: run `/squid-implement-night` (builds every pending task to a validated PR), or `/squid-implement-task` for individual tasks.

/squid-plan ends here.


Notes

  • Task-file shape: see the tracker-workflow spec (squid-scaffold/specs/tracker-workflow.md); ADR rules live in the product-architect agent contract.

© iusztinpaul, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/squid-plan of iusztinpaul/squid.

Open the folder on GitHubat commit f5bf6b3

Compare with similar skills

Squid Plan next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Squid Plan compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Squid Plan this skilliusztinpaul/squid203—~2.6kAutomated safety check: PassApache-2.0
Shep Workstreamsshep-ai/shep264—~2.5kAutomated safety check: PassMIT
Cabloy Spec Generationcabloy/cabloy982—~3.2kAutomated safety check: NotesMIT
Bmad Architectureaj-geddes/claude-code-bmad-skills488—~1.9kAutomated safety check: NotesCustom licence
Spec-Driven Developmentaddyosmani/agent-skills103k1 repos~3.2kAutomated safety check: PassMIT
MoAI SPEC Workflowmodu-ai/moai-adk1.2k—~5.1kAutomated safety check: PassApache-2.0

Similar skills

  • Shep Workstreams

    shep-ai/shep

    A skill your agent uses when a large body of work (a version milestone, an epic, a roadmap, a set of PRDs/design docs) needs to be broken into parallel workstreams and executed with the shep CLI.

    264 GitHub stars~2.5k tokensUpdated yesterday
    Product & Project ManagementAuto-check passed
  • A skill your agent uses to create or maintain Cabloy suite specifications under repo-specs, including PRD, SRS, PDP/WBS, acceptance planning, progress, and suite ADRs.

    982 GitHub stars~3.2k tokensUpdated today
    DevelopmentAuto-check: notes
  • Bmad Architecture

    aj-geddes/claude-code-bmad-skills

    Solutioning skill (Winston, the Architect). An agent skill from aj-geddes/claude-code-bmad-skills.

    488 GitHub stars~1.9k tokensUpdated 3 mo ago
    DevelopmentAuto-check: notes
  • Spec-Driven Development

    addyosmani/agent-skills

    Writes a structured specification before any code, moving through gated specify, plan, tasks and implement phases, with an optional capability map for multi-part requests.

    103k GitHub starsUsed in 1 repo~3.2k tokens
    DevelopmentAuto-check passed
  • MoAI SPEC Workflow

    modu-ai/moai-adk

    Manages SPEC documents for MoAI-ADK development, with GEARS or EARS requirement notation, acceptance criteria and a link into the Plan-Run-Sync workflow.

    1.2k GitHub stars~5.1k tokensUpdated today
    DevelopmentAuto-check passed
  • Foreman Grill Docs

    VisionForge-OU/foreman

    Headless grilling pass that challenges an approved implementation plan against the existing codebase and domain model, then writes an ADR draft and a PRD draft into the Foreman feature directory.

    443 GitHub stars~1.6k tokensUpdated 3 mo ago
    Agent WorkflowsAuto-check passed

More from iusztinpaul/squid

All 13 skills in this repo
  • Squid Implement Night

    iusztinpaul/squid

    Run the full agent-team pipeline end-to-end for one feature whose Tasks Plan is already approved by /squid-plan, handing the human a validated, ready-to-squash-merge PR.

    203 GitHub stars~1.2k tokensUpdated 1 mo ago
    Auto-check passed
  • Squid Implement Task

    iusztinpaul/squid

    Implement one task — or a whole list / an approved Tasks Plan — via the inner SWE↔Tester loop, committing each task on PASS.

    203 GitHub stars~1.3k tokensUpdated 1 mo ago
    Auto-check passed
  • Squid Review

    iusztinpaul/squid

    Push the committed feature branch, create or update its PR, then run Product Architect acceptance and PR-Reviewer on it.

    203 GitHub stars~873 tokensUpdated 1 mo ago
    Auto-check passed
  • Squid Review CI

    iusztinpaul/squid

    Drive CI green on a pushed, review-clean feature PR — On-Call diagnoses failures and hands fix tasks to the SWE.

    203 GitHub stars~607 tokensUpdated 1 mo ago
    Auto-check passed
  • Squid Testing Python

    iusztinpaul/squid

    Write and evaluate effective Python tests using pytest. An agent skill from iusztinpaul/squid.

    203 GitHub stars~1.3k tokensUpdated 1 mo ago
    Auto-check passed
  • Squid Architecture Review

    iusztinpaul/squid

    Periodic architectural sweep — reads existing ADRs, maps modules/dependencies/layering, and reports up to 10 prioritised findings shaped as refactor proposals /squid-refactor can consume directly.

    203 GitHub stars~2.6k tokensUpdated 1 mo ago
    Auto-check passed

Questions about Squid Plan

What does Squid Plan do?

Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides…. Squid Plan is an agent skill from iusztinpaul/squid. Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides everything touching the repo: tasks + storage, ADR, glossary, worktree, and which build to run.

When should I use Squid Plan?

Squid Plan fits situations like: tasks that involve Architecture decision records; tasks that involve PRD writing; tasks that involve Git worktrees.

How do I install Squid Plan in Claude Code?

Run `npx skills add iusztinpaul/squid --skill squid-plan -a claude-code`. Or copy the skill folder (skills/squid-plan in iusztinpaul/squid) into .claude/skills/squid-plan in your project. Claude Code loads it when a task matches its description.

How do I install Squid Plan in Codex?

Run `npx skills add iusztinpaul/squid --skill squid-plan -a codex`. Or copy the skill folder (skills/squid-plan in iusztinpaul/squid) into .agents/skills/squid-plan in your project. Codex loads it when a task matches its description.

Can I use Squid Plan in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add iusztinpaul/squid --skill squid-plan -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/squid-plan, .gemini/skills/squid-plan, .github/skills/squid-plan and .opencode/skills/squid-plan in your project.

What does Squid Plan need to run?

Going by SKILL.md and its folder, Squid Plan needs the command-line tools its instructions call (git and gh).

Does Squid Plan access the network?

SKILL.md contains no URLs. Its commands use git and gh, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Squid Plan safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Squid Plan use?

Squid Plan is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Squid Plan use?

About 2.6k tokens (SKILL.md is roughly 10k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Squid Plan?

Skills that share tags, products or a category with Squid Plan: Shep Workstreams (shep-ai/shep, 264 stars), Cabloy Spec Generation (cabloy/cabloy, 982 stars), Bmad Architecture (aj-geddes/claude-code-bmad-skills, 488 stars) and Spec-Driven Development (addyosmani/agent-skills, 103k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Squid Plan?

iusztinpaul (a GitHub user) maintains it in iusztinpaul/squid, which has 203 GitHub stars. The repository holds 13 skills in this directory. The repository was last updated on September 3, 2026.

Source: iusztinpaul/squid on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.