Scaffold an unconfigured directory into a configured pi project.

MITAuto-check passedAgent Workflows

Install Project Init

skills CLI
$ npx skills add BlackBeltTechnology/pi-agent-dashboard --skill project-init -a claude-code

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

GitHub CLI
$ gh skill install BlackBeltTechnology/pi-agent-dashboard project-init --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/BlackBeltTechnology/pi-agent-dashboard.git skills-src && mkdir -p .claude/skills && cp -r skills-src/packages/extension/.pi/skills/project-init .claude/skills/project-init && 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
project-init
GitHub stars
315
Token cost
~3.4k tokens
SKILL.md length
1,609 words
Files
11
Skills in repo
66
Repo updated
First seen
Licence
MIT

At a glance

Scaffold an unconfigured directory into a configured pi project.

  • Works in 7 steps: Resolve the available profiles → Ask which profile → Ask about optional capabilities,… → …
  • Tasks that involve Knowledge bases
  • SKILL.md covers Step 1 — Resolve the available…, Step 2 — Ask which profile, Step 2b — Resolve the… and Step 3 — Ask about optional…, plus 5 more sections
  • Calls npm, pnpm and yarn

What it does

Project Init is an agent skill from BlackBeltTechnology/pi-agent-dashboard. Scaffold an unconfigured directory into a configured pi project. Interactive, profile-driven: previews the planned writes, then writes AGENTS.md, .pi/settings.json and prompt files — optionally also a knowledge base, an openspec/ scaffold, and user-global ~/.pi/agent/settings.json. Use on a bare directory, or when the dashboard's "Initialize" button spawns this session.

Its SKILL.md is about 3.4k tokens, which your agent loads only when the skill is triggered. The skill folder holds 15 other files (for example `SKILL.md.AGENTS.md`, `profiles/coding/profile.json` and `profiles/coding/prompts/plan-change.md`).

It sits in Agent Workflows, covering Knowledge bases and Agent instruction files. It works with pnpm. The repository describes itself as: Real-time web dashboard for pi coding-agent sessions. Multi-session view, live chat mirroring, integrated terminal, diff viewer, pi-flows execution, and mobile-first remote… The licence is MIT.

When your agent uses it

  • Tasks that involve Knowledge bases
  • Tasks that involve Agent instruction files

Example prompts

  • “Initialize”
  • “/project-init”

Requirements

  • Python 3

Workflow steps

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

  1. Resolve the available profiles
  2. Ask which profile
  3. Ask about optional capabilities, preview, then confirm
  4. Write the scaffold
  5. DOX pointer seed (only when DOX_ENABLED is true)
  6. Run OpenSpec init (only when OPENSPEC_INIT is true)
  7. Done

What it can do on your machine

Read from SKILL.md and the folder at commit e23e533. 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:

    • npm
    • pnpm
    • yarn
    • bun
    • cargo
    • go
    • poetry
    • mvn
    • python

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

  • Network

    No URLs in SKILL.md. Its commands use npm, pnpm and yarn, 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

Project Init loads about 3.4k tokens when it runs. Until then it costs about 96 tokens; SKILL.md has 1,609 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~96
When it runs · the whole SKILL.md, loaded when a task matches
~3.4k

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 BlackBeltTechnology/pi-agent-dashboard at commit e23e533, republished under its MIT licence (© BlackBeltTechnology). 1,609 words, ~3,351 tokens.

Download SKILL.mdSave it as .claude/skills/project-init/SKILL.md (or your agent's skills folder). This skill also uses 10 other files; get the full folder from GitHub.
name
project-init
description
Scaffold an unconfigured directory into a configured pi project. Interactive, profile-driven: previews the planned writes, then writes AGENTS.md, .pi/settings.json and prompt files — optionally also a knowledge base, an openspec/ scaffold, and user-global ~/.pi/agent/settings.json. Use on a bare directory, or when the dashboard's "Initialize" button spawns this session.
license
MIT
metadata.author
pi-dashboard
metadata.version
1.0

Project Init

Turn the current directory into a configured pi project through a guided, conversational flow. You are running as a first-class interactive session — converse, confirm, and only write files after the user agrees.

Target directory = your current working directory (cwd). All writes land there. All profile/doctrine paths below are relative to THIS skill's directory (the folder containing this SKILL.md); call it <skill>.

Step 1 — Resolve the available profiles

A profile is a directory bundle: profile.json (optional, carries description + dox), AGENTS.md.tmpl, settings.json.tmpl, and prompts/*.md.

Enumerate profiles from two sources, user wins by name:

  1. Shipped: <skill>/profiles/*
  2. User: ~/.pi/project-profiles/*
bash
ls -1 "<skill>/profiles" 2>/dev/null
ls -1 "$HOME/.pi/project-profiles" 2>/dev/null

Merge: a user profile of the same name fully shadows the shipped one. A directory is only a usable profile when it contains BOTH AGENTS.md.tmpl and settings.json.tmpl. Read each profile.json for its description and dox flag (dox defaults to false when absent or unreadable).

Step 2 — Ask which profile

Use ask_user (select) to present the resolved profiles by name + description. Shipped defaults are coding (TDD/simplicity/surgical, OpenSpec on, npm ci worktree-init) and docs (writing structure, OpenSpec off, no build step).

Step 2b — Resolve the technology stack (stack-aware profiles only)

When the chosen profile's profile.json has "stackAware": true (the shipped coding profile does), its templates carry technology placeholders ({{INSTALL_CMD}}, {{TEST_CMD}}, {{BUILD_CMD}}, {{INIT_GATE}}, {{INIT_COMMAND}}) that must be filled from the project's stack — a coding repo may be Node, Rust, Go, Python, Java, etc.

Auto-detect + confirm:

  1. Detect a best guess from marker files in the target directory:

    Marker file(s)Stackinstall / test / buildinit gate + run
    pnpm-lock.yamlpnpmpnpm install --frozen-lockfile / pnpm test / pnpm buildtest ! -d node_modules / pnpm install --frozen-lockfile
    yarn.lockyarnyarn install --frozen-lockfile / yarn test / yarn buildtest ! -d node_modules / yarn install --frozen-lockfile
    bun.lockb / bun.lockbunbun install / bun test / bun run buildtest ! -d node_modules / bun install
    package-lock.json / package.jsonnpmnpm ci / npm test / npm run buildtest ! -d node_modules / npm ci
    Cargo.tomlcargocargo fetch / cargo test / cargo buildtest ! -d target / cargo fetch
    go.modgogo mod download / go test ./... / go build ./...test ! -f go.sum / go mod download
    poetry.lock or pyproject.toml+[tool.poetry]poetrypoetry install / poetry run pytest / poetry buildtest ! -d .venv / poetry install
    requirements.txt / pyproject.tomlpipvenv+pip install / .venv/bin/pytest / python -m buildtest ! -d .venv / venv+pip install
    pom.xmlmavenmvn -q dependency:go-offline / mvn test / mvn -q package -DskipTeststest ! -d target / mvn -q dependency:go-offline
    build.gradle(.kts)gradle./gradlew dependencies / ./gradlew test / ./gradlew build -x testtest ! -d .gradle / ./gradlew dependencies

    Prefer a JS lockfile over a plain package.json. A bare directory yields no guess.

  2. Confirm with the user (ask_user): show the detected stack (or "none detected") and let them confirm or pick a different stack from the table. Never scaffold a stack-aware profile without a resolved stack.

  3. Substitute the chosen stack's values for the {{…}} placeholders when writing the templates in Step 4. After writing, verify NO {{…}} placeholder remains in AGENTS.md or .pi/settings.json — an unfilled placeholder means the stack was not resolved; go back and ask.

Step 3 — Ask about optional capabilities, preview, then confirm

3a — Ask about DOX (all profiles)

DOX seeds a directory-level documentation doctrine and wires the kb toolset (see Step 5). It is now an interactive choice on every profile, not a fixed per-profile switch. Ask with ask_user (confirm), pre-selecting the profile's dox value as the DEFAULT answer:

Enable DOX documentation doctrine? Seeds a directory-level AGENTS.md discipline into ./AGENTS.md and enables the kb toolset. (default: <profile.dox>)

Record the answer as DOX_ENABLED. The dox flag in profile.json only sets the default; the user's answer decides. Step 5 gates on DOX_ENABLED, not the raw flag.

3b — Ask about OpenSpec init (coding profile only)

Gate: only when the selected profile is coding (skip for docs / any OpenSpec-off profile), mirroring Steps 4b and 5. Skip the prompt entirely when an openspec/ directory already exists in the target (idempotent — treat as already-initialized). Otherwise ask with ask_user (confirm):

Initialize OpenSpec for this project? Runs openspec init --tools pi, which creates the openspec/ scaffold and wires the OpenSpec commands into pi.

Record the answer as OPENSPEC_INIT. The actual run happens in Step 6.

3c — Preview + confirm

List exactly what you will write, then ask the user to confirm (ask_user confirm). The writes for profile <p> are:

  • ./AGENTS.md ← <skill>/profiles/<p>/AGENTS.md.tmpl (substitute {{PROJECT_NAME}} with the directory basename)
  • ./.pi/settings.json ← <skill>/profiles/<p>/settings.json.tmpl
  • ./.pi/prompts/*.md ← each file in <skill>/profiles/<p>/prompts/

When DOX_ENABLED is true, ALSO name in the preview:

  • the DOX pointer block appended to ./AGENTS.md (see Step 5)
  • the doctrine choice + kb toolset flip written to ./.pi/dashboard/knowledge_base.json

When OPENSPEC_INIT is true, ALSO disclose the side effect (see Step 6) — NOT a plain file write:

  • runs openspec init --tools pi, scaffolding ./openspec/ and wiring pi

When the chosen profile is coding, ALSO disclose the possible side effect (see Step 4b) — NOT a file write:

  • may offer to pi install the eng-disciplines skills user-globally (writes ~/.pi/agent/settings.json). This is machine-wide: the skills become available in all projects on this machine, not just this one. Always opt-in via a separate prompt; never forced. Needs Node/npm on the machine (the install runs pi install npm:…); the scaffolded repo itself gains no dependency.

Idempotency: before writing, check whether ./AGENTS.md or ./.pi/settings.json already exist. If they do, ask before overwriting — never clobber silently.

Step 4 — Write the scaffold

On confirmation, write the files listed above. Substitute {{PROJECT_NAME}} in AGENTS.md.tmpl with the target directory's basename, and — for a stack-aware profile — the stack placeholders resolved in Step 2b.

Validate the hook. After writing ./.pi/settings.json, confirm its worktreeInit is a valid change-A hook: a non-empty gate string plus a run that is either { type: "script", command: "<non-empty>" } or { type: "agent", prompt: "<non-empty>" }. If it is not valid, warn the user — an invalid hook fails open (change-A ignores it) and the Initialize button will loop back to this skill instead of running the hook.

Show full SKILL.md (655 more words)Show less

Step 4b — Ensure discipline skills (only when the profile is coding)

The coding template's ## Discipline Skills checkpoint table references the eng-disciplines skills. This step makes those references live. Gate: run only when the selected profile is coding (skip for docs / any OpenSpec-off profile), mirroring how Step 5 gates on dox: true.

Detect (read-only; both forms tolerate a missing pi binary):

bash
pi list 2>/dev/null | grep -q pi-dashboard-eng-disciplines \
  || stat ~/.pi/agent/npm/node_modules/@blackbelt-technology/pi-dashboard-eng-disciplines >/dev/null 2>&1

If pi is not on PATH and the stat misses, treat as ABSENT but do NOT error the init — fall through to the footnote. (Known limit: the stat form only sees the npm: global path; a git:/renamed install is invisible and would re-prompt.)

PRESENT → skills already global. Skip the prompt (idempotent re-run) and write NO activation footnote into ./AGENTS.md. If a prior run left the "not detected" footnote (e.g. the skills were installed after a decline), remove that line before exiting so the file never claims the skills are missing when they are present. Done.

ABSENT → ask_user (confirm):

Install the discipline skills globally? They power the checkpoint table this project's AGENTS.md references, and become available in ALL projects on this machine. Writes ~/.pi/agent/settings.json (user-global, not project-local).

  • Yes → run pi install npm:@blackbelt-technology/pi-dashboard-eng-disciplines and verify exit 0. On success, write NO footnote. On non-zero exit (or a missing pi), fall through to the footnote path.

  • No / install failed → append ONE line under the ## Discipline Skills table in ./AGENTS.md:

    Discipline skills not detected — run pi install npm:@blackbelt-technology/pi-dashboard-eng-disciplines to activate the checkpoints above.

The footnote is detection-conditional: written only on the absent/declined branch, never on the present/installed path (so a successfully-wired project never carries a false "not detected" line). The install is always user-global and never forced.

Step 5 — DOX pointer seed (only when DOX_ENABLED is true)

Gated on the user's Step 3a answer (DOX_ENABLED), NOT the raw profile.json dox flag — the flag only supplied the default. Skip this step entirely when the user declined DOX.

Do NOT copy any doctrine text. The per-turn DOX doctrine (READ + WRITE) is injected by the pi-dashboard-kb-extension; the project's AGENTS.md carries only a marker + pointer block. Tune it via .pi/dashboard/knowledge_base.json.

1. Seed the pointer block. Append it to ./AGENTS.md only when it does not already carry the marker <!-- dox-doctrine --> (idempotent — if the marker is present, skip this step entirely):

markdown
<!-- dox-doctrine -->

## DOX doctrine

Per-turn DOX doctrine — the kb-first READ discipline and the directory `AGENTS.md`
WRITE discipline — is injected by the `pi-dashboard-kb-extension`. Tune it in
`.pi/dashboard/knowledge_base.json` under the `doctrine` key (`inject`, `write`).

2. Record the doctrine choice. Read-merge-write ./.pi/dashboard/knowledge_base.json — read the file if present, set ONLY the doctrine key, preserve every other key, write valid JSON. Since DOX_ENABLED is true here, write:

json
"doctrine": { "inject": "kb", "write": true }

When the file is absent, also enable the directory-level AGENTS.md toolset:

json
{
  "sources": [{ "kind": "filesystem", "ref": "." }],
  "indexAgentsFiles": true,
  "directoryLevelAgents": { "enabled": true },
  "doctrine": { "inject": "kb", "write": true }
}

3. Scaffold the directory AGENTS.md tree (path-only rows):

bash
kb dox init

4. Offer dox-describe. After the tree is scaffolded, ask the user (ask_user confirm): "Fill the empty Purpose cells now with dox-describe? (one subagent per AGENTS.md, plan-then-confirm, ≤50 rows per run)". On yes, invoke the dox-describe skill (ships with packages/kb-extension); on no, tell the user they can run it later.

Step 6 — Run OpenSpec init (only when OPENSPEC_INIT is true)

Gated on the user's Step 3b answer (OPENSPEC_INIT); this only becomes true for the coding profile when no openspec/ dir pre-existed. Mirrors Step 4b's side-effect pattern — opt-in, verified, never forced.

Run (always non-interactive so it never hijacks the conversation):

bash
openspec init --tools pi

The --tools pi flag pins the wiring to pi and skips OpenSpec's interactive tool picker. Verify exit 0 and that ./openspec/ now exists. On success, tell the user OpenSpec is wired (the coding AGENTS.md already documents openspec change new <name>). On non-zero exit (or a missing openspec binary), warn the user and point them at openspec init --tools pi to run manually — do NOT error the init; the scaffold is already written.

Idempotent re-run: if ./openspec/ already existed, Step 3b never asked, so OPENSPEC_INIT is false and this step is skipped — no clobber.

Step 7 — Done

Confirm what was written. Writing worktreeInit flips the directory to "configured": the next dashboard Initialize click runs the hook (change-A) instead of re-launching this skill. Tell the user they can click Initialize again to run npm ci (or the profile's hook).

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

Files

SKILL.md and 10 other files in packages/extension/.pi/skills/project-init of BlackBeltTechnology/pi-agent-dashboard.

  • SKILL.md
  • SKILL.md.AGENTS.md
  • profiles/coding/AGENTS.md.tmpl
  • profiles/coding/profile.json
  • profiles/coding/prompts/plan-change.md
  • profiles/coding/prompts/tdd-loop.md
  • profiles/coding/settings.json.tmpl
  • profiles/docs/AGENTS.md.tmpl
  • profiles/docs/profile.json
  • profiles/docs/prompts/draft-page.md
  • profiles/docs/settings.json.tmpl

Open the folder on GitHubat commit e23e533

Compare with similar skills

Project Init 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.

Project Init compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Project Init this skillBlackBeltTechnology/pi-agent-dashboard315—~3.4kAutomated safety check: PassMIT
Docs Gardeningmhmzdev/the-holy-quran-app889—~1.1kAutomated safety check: PassMIT
Sake Workflowkattouf/Sake116—~935Automated safety check: PassMIT
Claude Code Knowledgedykyi-roman/awesome-claude-code103—~4kAutomated safety check: PassMIT
Agent Doc Sync Toollittleben/awesomeAgentskills1882 repos~470Automated safety check: PassMIT
Promotealirezarezvani/claude-skills28k2 repos~1.1kAutomated safety check: PassMIT

Similar skills

  • Docs Gardening

    mhmzdev/the-holy-quran-app

    Audit and maintain The Holy Qur'an app's knowledge base (AGENTS.md, .agents/rules/, docs/ OKF bundle) for staleness, broken cross-links, drift from code, and OKF structural issues.

    889 GitHub stars~1.1k tokensUpdated 28 days ago
    Agent WorkflowsAuto-check passed
  • Sake Workflow

    kattouf/Sake

    A skill your agent uses when starting any new task in the Sake project - ensures structured execution with knowledge discovery, quality checks, and organic growth of the project knowledge base…

    116 GitHub stars~935 tokensUpdated 6 mo ago
    Agent WorkflowsAuto-check passed
  • Claude Code Knowledge

    dykyi-roman/awesome-claude-code

    Comprehensive knowledge base for Claude Code formats, patterns, and configuration.

    103 GitHub stars~4k tokensUpdated 1 mo ago
    Agent WorkflowsAuto-check passed
  • Agent Doc Sync Tool

    littleben/awesomeAgentskills

    Keeps Agents.md, claude.md and gemini.md identical across a project by scanning folders and copying the newest version over the others, once or continuously.

    188 GitHub starsUsed in 2 repos~470 tokens
    Agent WorkflowsAuto-check passed
  • Promote

    alirezarezvani/claude-skills

    Graduate a proven pattern from auto-memory (MEMORY.md) to CLAUDE.md or .claude/rules/ for permanent enforcement.

    28k GitHub starsUsed in 2 repos~1.1k tokens
    Agent WorkflowsAuto-check passed
  • Project Memory

    jamditis/claude-skills-journalism

    Generates CLAUDE.md project memory files that transfer institutional knowledge.

    416 GitHub stars~2.5k tokensUpdated 3 days ago
    Agent WorkflowsAuto-check: notes

More from BlackBeltTechnology/pi-agent-dashboard

All 66 skills in this repo
  • Browser

    BlackBeltTechnology/pi-agent-dashboard

    Browser automation via the agent-browser CLI. An agent skill from BlackBeltTechnology/pi-agent-dashboard.

    315 GitHub starsUsed in 1 repo~2k tokens
    Auto-check passed
  • Pi Dashboard

    BlackBeltTechnology/pi-agent-dashboard

    Monitor and control the pi-dashboard server. An agent skill from BlackBeltTechnology/pi-agent-dashboard.

    315 GitHub starsUsed in 1 repo~2.2k tokens
    Auto-check passed
  • CI Troubleshoot

    BlackBeltTechnology/pi-agent-dashboard

    Diagnose failed GitHub Actions runs for pi-agent-dashboard: the 11-file workflow taxonomy, affected-test selection, the release pipeline, known failure modes, and how to read gh run logs and…

    315 GitHub stars~3.5k tokensUpdated today
    Auto-check passed
  • Debug Dashboard

    BlackBeltTechnology/pi-agent-dashboard

    Diagnose problems in the running pi-agent-dashboard system: server.log, /api/health, bridge WebSocket connectivity, vitest triage, known-issue FAQ entries.

    315 GitHub stars~1.6k tokensUpdated today
    Auto-check passed
  • Implement

    BlackBeltTechnology/pi-agent-dashboard

    Disciplined implementation in pi-agent-dashboard: the rebuild matrix (extension→reload, server→restart, client→build+restart, openspec-apply→full rebuild) plus the project's code discipline rules.

    315 GitHub stars~2k tokensUpdated today
    Auto-check passed
  • Session To Guideline

    BlackBeltTechnology/pi-agent-dashboard

    Turn a pi session into a Markdown "how-we-did-it" collaboration guideline: reads the session's JSONL transcript and synthesizes a reusable playbook of which prompts worked, what had to be steered…

    315 GitHub stars~3.2k tokensUpdated today
    Auto-check passed

Works with

Questions about Project Init

What does Project Init do?

Scaffold an unconfigured directory into a configured pi project. Project Init is an agent skill from BlackBeltTechnology/pi-agent-dashboard. Scaffold an unconfigured directory into a configured pi project.

When should I use Project Init?

Project Init fits situations like: tasks that involve Knowledge bases; tasks that involve Agent instruction files.

How do I install Project Init in Claude Code?

Run `npx skills add BlackBeltTechnology/pi-agent-dashboard --skill project-init -a claude-code`. Or copy the skill folder (packages/extension/.pi/skills/project-init in BlackBeltTechnology/pi-agent-dashboard) into .claude/skills/project-init in your project. Claude Code loads it when a task matches its description.

How do I install Project Init in Codex?

Run `npx skills add BlackBeltTechnology/pi-agent-dashboard --skill project-init -a codex`. Or copy the skill folder (packages/extension/.pi/skills/project-init in BlackBeltTechnology/pi-agent-dashboard) into .agents/skills/project-init in your project. Codex loads it when a task matches its description.

Can I use Project Init 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 BlackBeltTechnology/pi-agent-dashboard --skill project-init -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/project-init, .gemini/skills/project-init, .github/skills/project-init and .opencode/skills/project-init in your project.

What does Project Init need to run?

Going by SKILL.md and its folder, Project Init needs the command-line tools its instructions call (npm, pnpm, yarn, bun, cargo and go). Our summary lists: Python 3.

Does Project Init access the network?

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

Is Project Init 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 Project Init use?

Project Init is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Project Init use?

About 3.4k tokens (SKILL.md is roughly 13k 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 Project Init?

Skills that share tags, products or a category with Project Init: Docs Gardening (mhmzdev/the-holy-quran-app, 889 stars), Sake Workflow (kattouf/Sake, 116 stars), Claude Code Knowledge (dykyi-roman/awesome-claude-code, 103 stars) and Agent Doc Sync Tool (littleben/awesomeAgentskills, 188 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Project Init?

BlackBeltTechnology (a GitHub organization) maintains it in BlackBeltTechnology/pi-agent-dashboard, which has 315 GitHub stars. The repository holds 66 skills in this directory. The repository was last updated on October 8, 2026.

Source: BlackBeltTechnology/pi-agent-dashboard on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.