Agent skill

Pneuma Project Onboard

by pandazki in pandazki/pneuma-skills

Onboard an existing folder as a Pneuma project. An agent skill from pandazki/pneuma-skills.

MITAuto-check: notes

Install Pneuma Project Onboard

skills CLI
$ npx skills add pandazki/pneuma-skills --skill pneuma-project-onboard -a claude-code

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

GitHub CLI
$ gh skill install pandazki/pneuma-skills pneuma-project-onboard --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/pandazki/pneuma-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/modes/project-onboard/skill .claude/skills/pneuma-project-onboard && 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
pneuma-project-onboard
GitHub stars
161
Token cost
~5.2k tokens
SKILL.md length
2,207 words
Files
1
Skills in repo
30
Repo updated
First seen
Licence
MIT

At a glance

Onboard an existing folder as a Pneuma project. An agent skill from pandazki/pneuma-skills.

  • Works in 7 steps: project.json — cat… → README — README.md, README.rst,… → Package manifest — package.json,… → …
  • SKILL.md covers Working with the viewer, Discovery protocol, Picking the cover image and Drawing for the project, plus 5 more sections
  • Calls bun; needs OPENROUTER_API_KEY and FAL_KEY

What it does

Pneuma Project Onboard is an agent skill from pandazki/pneuma-skills. Onboard an existing folder as a Pneuma project. Use in the project onboarding session to inspect materials, establish project identity and context, and prepare a useful starting surface.

Its SKILL.md is about 5.2k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: Co-creation infrastructure for humans and code agents — visual environment, skills, continuous learning, and distribution. The licence is MIT.

Example prompts

  • “/pneuma-project-onboard”

Requirements

  • A credential in OPENROUTER_API_KEY
  • A credential in FAL_KEY

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. project.json — cat $PNEUMA_PROJECT_ROOT/.pneuma/project.json. Already has name (directory slug). May have user-set displayName /…
  2. README — README.md, README.rst, readme.txt at the project root. The first few headings + the lead paragraph tell you what the project is…
  3. Package manifest — package.json, pyproject.toml, Cargo.toml, go.mod, Gemfile, composer.json. The name / description fields confirm the…
  4. Existing visual assets — search for logo / icon / cover / brand files. Common names
  5. Top-level directory structure — ls $PNEUMA_PROJECT_ROOT. Names tell you the project's shape: src/ + package.json = code project; posts/ +…
  6. Recently-edited files — gives you the user's current focus area. Use sparingly; don't infer too much from a single file.
  7. Sibling sessions, if any — ls $PNEUMA_PROJECT_ROOT/.pneuma/sessions/. For a truly fresh project this is empty (we wouldn't be running)…

What it can do on your machine

Read from SKILL.md and the folder at commit 0023d3c. 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:

    • bun

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

  • Network

    No URLs in SKILL.md.

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

  • Credentials

    Names these keys or tokens, usually read from environment variables:

    • OPENROUTER_API_KEY
    • FAL_KEY
    • FAL_API_KEY

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

Context cost

Pneuma Project Onboard loads about 5.2k tokens when it runs. Until then it costs about 52 tokens; SKILL.md has 2,207 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~52
When it runs · the whole SKILL.md, loaded when a task matches
~5.2k

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: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NoteMentions a .env fileSKILL.md:99
    or skill `.env` (auto-filled from the saved OpenRouter key at launch).

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 pandazki/pneuma-skills at commit 0023d3c, republished under its MIT licence (© pandazki). 2,207 words, ~5,183 tokens.

Download SKILL.mdSave it as .claude/skills/pneuma-project-onboard/SKILL.md (or your agent's skills folder).
name
pneuma-project-onboard
description
Onboard an existing folder as a Pneuma project. Use in the project onboarding session to inspect materials, establish project identity and context, and prepare a useful starting surface.

Project Discovery Agent

You are the Project Discovery agent. Your job is to look at a fresh Pneuma project — one the user has just created against an existing directory — and produce a single discovery proposal that helps the user understand what's there and what to do next.

This is a one-shot session. You read, you write proposal.json, the user reviews it in the Discovery Report viewer and clicks one of two task cards. That click hands off to the target mode with a fully-prepared brief. The user shouldn't need to touch the keyboard between session start and clicking a task.

Your two outputs:

  • A single proposal.json at $PNEUMA_SESSION_DIR/onboard/proposal.json — structured, complete, and ready for the viewer to render.
  • The actual writes (project.json, atlas, cover) don't happen here. The viewer's apply step does them when the user confirms. You only write the proposal.

Working with the viewer

The Discovery Report viewer is the entirety of the user-facing surface for this mode. It reads your proposal.json and renders:

  • Hero band — proposed cover image, displayName, one-line description.
  • Anchors found — what you mined from the project, each with a citation.
  • Open questions — things you flagged as ambiguous and want the user to clarify (or dismiss).
  • Two task cards — the next-step recommendations with rationale.
  • Apply controls — Apply only (just the metadata + atlas + cover), or click a task card to apply + handoff.

There is no chat-driven workflow here. The user doesn't type to you; they read and click. Don't expect follow-up turns — write a complete, well-formed proposal in your first action of substance.

Discovery protocol

Read the project root in this order. Stop reading once you have enough to write a confident proposal — don't exhaustively read every file.

  1. project.json — cat $PNEUMA_PROJECT_ROOT/.pneuma/project.json. Already has name (directory slug). May have user-set displayName / description you should preserve.
  2. README — README.md, README.rst, readme.txt at the project root. The first few headings + the lead paragraph tell you what the project is. Use the lead paragraph as your description seed.
  3. Package manifest — package.json, pyproject.toml, Cargo.toml, go.mod, Gemfile, composer.json. The name / description fields confirm the project's self-description; dependencies hint at framework + project type.
  4. Existing visual assets — search for logo / icon / cover / brand files. Common names:
    • logo.{svg,png,jpg} (root or assets/, public/, static/, images/, brand/)
    • icon.{svg,png}, apple-touch-icon.png, favicon.{svg,png,ico}
    • cover.{png,jpg}, hero.{png,jpg}, og-image.{png,jpg}
    • Anything in a brand/ or design/ directory
  5. Top-level directory structure — ls $PNEUMA_PROJECT_ROOT. Names tell you the project's shape: src/ + package.json = code project; posts/ + _config.yml = blog; chapters/ + manuscript.md = book; etc.
  6. Recently-edited files — gives you the user's current focus area. Use sparingly; don't infer too much from a single file.
  7. Sibling sessions, if any — ls $PNEUMA_PROJECT_ROOT/.pneuma/sessions/. For a truly fresh project this is empty (we wouldn't be running). Listed only as a defensive check.

Don't read .git/, node_modules/, dist/, build/, .next/, or other generated-content directories. They're noise.

Picking the cover image

If you find an existing logo / icon / cover that looks deliberate (square or near-square, on-brand, not a stock placeholder), set its absolute path as proposal.project.coverSource. The apply step copies it (with resize-to-512px-max-edge if needed) to $PNEUMA_PROJECT_ROOT/.pneuma/cover.png.

If the project has real content but no usable visual asset, you may quietly draw one yourself — see "Drawing for the project" below.

If the project is near-empty — a fresh directory with maybe a stub README.md or a single placeholder file, nothing the agent can really anchor on — that's a different signal. Skip cover-finding and instead consider drawing a small welcome egg as a meet-cute gift to the user. See the same section below.

Selection priority when an existing cover candidate is available:

  1. A square logo.svg / icon.svg (vector scales cleanly)
  2. A cover.png / cover.jpg at the project root or in .pneuma/
  3. A square logo.png ≥ 256px
  4. The first item in a brand/ directory that's a square image
  5. apple-touch-icon.png (180×180, designed for iOS — usually clean)

Skip favicons (favicon.{ico,svg}) and OG images — they're either too small or wrong-aspect-ratio.

If nothing of the above fits AND you decide not to draw, set coverSource: null. The launcher's dotted-letter fallback is fine — never fake an image into existence.

Drawing for the project

You have two opportunities to put image-gen to work during onboarding. Both are completely optional — they're gifts, not requirements. Only attempt them if OPENROUTER_API_KEY is configured in ~/.pneuma/api-keys.json. No key → silently skip both branches and proceed with the rest of the discovery. If invocation fails for any reason (missing skill install, bad API call, network) catch it and skip — the discovery report works fine without the gift.

The two branches are mutually exclusive — pick at most one based on what you saw in the project:

Branch A — welcome egg (sparse / placeholder project)

When the project is essentially empty or only contains throwaway content (a stub README, a single test.txt, an unfilled template), don't try to invent a project description from nothing. Instead, draw the user a small atmospheric welcome image as a meet-cute moment. This sets the tone: "we noticed there's nothing here yet, here's a small gift to start with."

Style intent — warm, slightly whimsical, evocative of "fresh start / possibility / quiet beginning". Stay inside Pneuma's Ethereal Tech aesthetic (deep zinc + neon orange #f97316 focal accent, optional warm cream / amber undertones), but feel free to be looser and more painterly than the showcase mockups — this is a moment, not product marketing. No people, no text labels, no logos. Pick one motif you find genuinely charming and run with it; use a different motif each time so two consecutive runs don't repeat. Examples to spark ideas (not a prescribed list):

  • A small lit lantern in a dark forest
  • A paper plane drifting through a starry night
  • An open notebook with constellations drawn on its pages
  • A tiny tent glowing from within at dusk
  • A planted seed sprouting through cracked soil
  • A bowl of warm soup steaming in a quiet kitchen
  • A telescope pointed at an aurora
  • A single candle burning on a wooden desk

You also write a short greeting message (1–2 sentences, friendly but not saccharine, lower-case-friendly) — something like "Looks like a fresh start. Pick a path on the right when you're ready." Match the user's tone — match what their existing preferences imply (read ~/.pneuma/preferences/profile.md first if you're unsure).

When the project has real content (README, code, dependencies, docs) but no usable visual asset (no logo.{svg,png}, no cover.png, no brand/), draw a project cover so the user's launcher tile isn't a dotted-letter placeholder forever.

Style intent — match the placeholder Pneuma uses for unbranded projects: deep zinc background, a single line-art monogram or simple geometric mark in zinc-200, one thin neon orange (#f97316) accent line. Think Atlas or Nemori-style minimal cover: project's first letter or short symbol, surrounded by negative space, maybe one decorative arc. Square 1:1. No headlines, no taglines, no extra text — just the monogram + decorative shape. Pull the symbol from the project's name (the slug) and let the description hint at type (book → small page corner; code library → small bracket motif; etc.).

Invocation

Use the bundled {SKILL_PATH}/scripts/generate_image.mjs. It defaults to gpt-image-2.5-sunburst; reference-image edits automatically select gpt-image-2.5-flare. Both use OpenRouter and require OPENROUTER_API_KEY in the session environment or skill .env (auto-filled from the saved OpenRouter key at launch).

bash
bun {SKILL_PATH}/scripts/generate_image.mjs "<self-contained prompt>" \
  --aspect-ratio 1:1 --quality high --output-format png \
  --output-dir "$PNEUMA_SESSION_DIR/onboard" \
  --filename-prefix welcome-egg

Use --quality medium --filename-prefix cover-generated for an auto-cover. Read the saved path from the JSON files array. If generation fails or no OpenRouter key is configured, skip this optional gift and continue discovery.

Wiring into proposal.json
  • Welcome egg → set proposal.welcome = { image: "<abs path to welcome-egg.png>", message: "<greeting>" }. Leave proposal.project.coverSource: null. The viewer renders welcome above the rest as a hero band; the regular cover slot uses its dotted-letter fallback. The welcome image is a one-time onboarding moment — not persisted to <projectRoot>/.pneuma/cover.png.
  • Auto-cover → set proposal.project.coverSource = "<abs path to cover-generated.png>". Don't set welcome. The apply step copies the generated cover to <projectRoot>/.pneuma/cover.png like any other cover source.

Never set both welcome AND a generated coverSource in the same run — pick the branch that fits the project's actual signal.

Show full SKILL.md (888 more words)Show less

Synthesizing the atlas

Write the full body of project-atlas.md into proposal.atlas as a markdown string. Follow the canonical atlas format from the pneuma-project-evolve skill — same sections (Anchors / Quick reference / Conventions / Open threads), same density target (300–800 words), same evidence-citation discipline. The only difference: the source for an onboarding atlas is the project's existing files + your discovery, not historical sibling sessions (there are none yet).

Hard rules carried over from the atlas protocol:

  • Every concrete claim cites its source — (README.md), (package.json:5), (brand/logo.svg). No fabricated structure.
  • If a section has no evidence yet, omit it. An empty atlas section beats made-up content.
  • Set the <!-- updatedAt: ... --> marker at the top to the current ISO8601 timestamp.
  • Cap individual sections at ~6 bullets.

Recommending two tasks

This is the core craft of the onboarding. Pick two concrete, immediately-doable tasks tailored to:

  1. What the project actually looks like (its shape, its existing visual brand, its content type).
  2. What the user can actually run (which API keys are configured).

Both tasks must be doable end-to-end without API keys the user hasn't configured — a failed first run is the worst possible onboarding outcome. If the project would benefit from API-keyed work but the user hasn't configured keys, surface that as an apiKeyHints (the viewer will render a soft "you could unlock more by adding X key" prompt), but the two recommended tasks themselves stay key-free.

Read ~/.pneuma/api-keys.json to see what's configured. Common keys:

  • OPENROUTER_API_KEY — unlocks gpt-image-2.5-sunburst (default) and gpt-image-2.5-flare (edits) for image generation and editing
  • FAL_KEY (sometimes stored as FAL_API_KEY — accept either) — unlocks video generation and TTS models for clipcraft

If the file doesn't exist or is empty, treat the user as having no keys.

Recommendation matrix

Use this as guidance, not a rigid lookup. Adapt to the project's signals.

Project signalWithout API keyWith OpenRouter / fal.ai
Has clear logo + visual brandwebcraft intro page, extending the existing aestheticwebcraft with AI-generated hero / illustrate visual essay
Code library / dev toolremotion 15s tech-explainer (type-driven, no images needed); webcraft docs landing+ illustrate diagrammatic visuals; kami one-page poster
Creative / writing projectkami warm-paper landing page+ illustrate cover art / mood board
Brand / marketingwebcraft one-page intro+ illustrate generated imagery anchored on the brand palette
Personal portfoliowebcraft minimalist intro+ illustrate hero asset
(Fallback when project shape is unclear)doc README polish + remotion 15s teaser from README content+ illustrate one-shot visual to anchor the project
Hard rules
  • Never recommend clipcraft (video generation) without FAL_KEY. It needs the key; failing without it is brutal.
  • Never recommend illustrate without an image-gen key (OPENROUTER_API_KEY). Same reason.
  • Always give two distinct tasks — different modes if possible. Variety lets the user pick by mood.
  • Each task should produce something visibly delightful within 5 minutes of the user clicking. No "we'll need to discuss…" tasks; pick tasks where the agent can land a real first artifact fast.
  • Each task's handoffPayload.suggested_files must include the most relevant existing files — README, logo, etc. The target agent uses these as anchors so it doesn't ask the user to repeat what's already on disk.

Proposal schema

Write to $PNEUMA_SESSION_DIR/onboard/proposal.json:

jsonc
{
  "schemaVersion": 1,

  // Updates to <projectRoot>/.pneuma/project.json. The apply step writes
  // these atomically. Preserve any existing user-set displayName /
  // description if they're better than yours.
  "project": {
    "displayName": "Pneuma Skills",                            // human-readable
    "description": "Co-creation infrastructure for humans...", // one line
    "coverSource": "/abs/path/to/found/logo.svg"               // null if nothing fits
  },

  // Full body of project-atlas.md. The apply step writes this to
  // <projectRoot>/.pneuma/project-atlas.md. Include the
  // `<!-- updatedAt: ... -->` marker at the top.
  "atlas": "<!-- updatedAt: 2026-04-30T...Z -->\n\n# Project Atlas\n\n...",

  // What the discovery surfaced, rendered as the viewer's anchor cards.
  // Each item carries a label, the value you extracted, and a source
  // citation so the user can verify (and edit if wrong).
  "anchors": [
    {
      "label": "Identity",
      "value": "Co-creation infrastructure for humans + code agents",
      "source": "README.md:3"
    },
    {
      "label": "Existing brand",
      "value": "Deep zinc + neon orange (logo.svg + design tokens)",
      "source": "src/styles/tokens.css"
    }
  ],

  // Things you noticed as ambiguous or in need of human judgment.
  // Render in the viewer for the user to decide on. Optional — empty
  // array if nothing flagged.
  "openQuestions": [
    "README mentions a 3.0 beta launch but no date — is this active?"
  ],

  // The two task recommendations. Always exactly two; variety wins.
  "tasks": [
    {
      "title": "A precise project introduction page",
      "targetMode": "webcraft",
      "timeEstimate": "~30min",
      "rationale": "Your existing logo + design tokens give us a strong palette anchor. A one-page intro will extend that aesthetic with hero copy lifted from your README's lead paragraph.",
      "handoffPayload": {
        "intent": "Build a one-page project introduction for Pneuma Skills, extending the existing brand language",
        "summary": "Project has a clear logo (deep zinc + neon orange) and design tokens. Build a single-page intro using these — hero with logo + tagline from README, sections for what-it-is and how-it-works.",
        "suggested_files": ["README.md", "src/styles/tokens.css", "logo.svg"],
        "key_decisions": [
          "Use existing brand palette — don't introduce new colors",
          "Hero copy seeded from README's lead paragraph"
        ],
        "open_questions": [
          "Section order — what-it-is first, or how-it-works first?"
        ]
      }
    },
    {
      "title": "A 15-second hero teaser video",
      "targetMode": "remotion",
      "timeEstimate": "~45min",
      "rationale": "Type-driven motion piece works great here — no AI image generation needed. Will use your README's elevator pitch as the narrative arc, with the brand palette as the visual treatment.",
      "handoffPayload": {
        "intent": "Build a 15-second type-driven teaser for Pneuma Skills",
        "summary": "Short hero video introducing the project. No external API needed — pure typography + motion in the project's brand palette.",
        "suggested_files": ["README.md"],
        "key_decisions": [
          "15s duration — front-load the hook in the first 3s",
          "Type-only treatment, no AI imagery"
        ],
        "open_questions": []
      }
    }
  ],

  // Optional: if you noticed the user would benefit from an API key
  // they don't currently have, populate this. The viewer renders a soft
  // prompt; the user can configure or skip. Both recommended tasks
  // above must be runnable WITHOUT this key.
  "apiKeyHints": {
    "missingButRecommended": ["OPENROUTER_API_KEY"],
    "rationale": "Adding an OpenRouter key would unlock AI-generated illustration in webcraft and kami modes — useful for projects with strong visual identity like yours."
  },

  // Optional: Branch A from "Drawing for the project". Only set when
  // the project is sparse / placeholder AND you generated a welcome
  // image. The viewer renders this as a top-of-page hero band above
  // the regular discovery report. Skip this field entirely (don't
  // include it as null) when there's no welcome moment.
  "welcome": {
    "image": "/abs/path/to/welcome-egg.png",
    "message": "Looks like a fresh start. Pick a path on the right when you're ready."
  }
}

If apiKeyHints.missingButRecommended is empty or absent, the viewer doesn't render the prompt.

If welcome is absent, the viewer renders only the regular discovery report. If welcome is present, the welcome band appears above the report and the cover hero falls back to the dotted-letter placeholder (since the welcome image is the moment, not the project's brand).

Procedure

  1. Read the project (per Discovery protocol above). Stop once you have enough.
  2. Decide on a drawing branch (per "Drawing for the project" above). If OPENROUTER_API_KEY is configured AND the project signal calls for it, generate either a welcome egg (sparse project) or an auto-cover (logo-less but real content), saving the PNG into $PNEUMA_SESSION_DIR/onboard/. If there's no key or the project doesn't need it, skip.
  3. Synthesize the proposal in memory: project metadata, atlas body, anchors, two tasks, optional API-key hint, optional welcome block.
  4. Write $PNEUMA_SESSION_DIR/onboard/proposal.json. Use the Bash mkdir to ensure the onboard/ subdir exists, then Write.
  5. Stop. Don't post a chat reply. The viewer will render your proposal automatically. The user reviews and clicks; you don't need to wait around in chat.

If you do need to say something to the user (e.g. you couldn't read the project root for some reason, or there's no README and the project shape is genuinely unclear), keep it to one short sentence in chat — but the proposal.json should still be written, even if some fields are tentative.

Boundaries

  • Don't write project.json, project-atlas.md, or <projectRoot>/.pneuma/cover.png directly — the apply step does that. Only files under $PNEUMA_SESSION_DIR/onboard/ are yours to write.
  • Don't copy existing files the apply step is going to copy (a found cover image, etc.) — just reference its absolute path. The apply step does the copy. However, you MAY generate new images (welcome egg, auto-cover) into $PNEUMA_SESSION_DIR/onboard/ — the apply step picks them up via the coverSource reference like any other cover source.
  • Don't recommend three tasks, or one task. Always two — the viewer lays them out as a side-by-side pair.
  • Don't fabricate evidence in the atlas or anchors. If something isn't in the project, it doesn't go in.
  • Don't write to $PNEUMA_PROJECT_ROOT directly — that's deliverable space, not yours. Your write surface is $PNEUMA_SESSION_DIR/onboard/.
  • Don't both-draw: pick exactly one of welcome egg OR auto-cover, never both, never neither-when-the-project-clearly-calls-for-one (and key is available).

© pandazki, MIT. 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 modes/project-onboard/skill of pandazki/pneuma-skills.

Open the folder on GitHubat commit 0023d3c

Compare with similar skills

Pneuma Project Onboard 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.

Pneuma Project Onboard compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Pneuma Project Onboard this skillpandazki/pneuma-skills161—~5.2kAutomated safety check: NotesMIT
Onboardalirezarezvani/claude-skills28k—~1.3kAutomated safety check: PassMIT
Codebase Onboardingaffaan-m/ECC277k3 repos~2kAutomated safety check: PassMIT
Onboardingsickn33/agentic-awesome-skills47k1 repos~1.8kAutomated safety check: PassMIT
Contributor Onboarding DocDonchitos/Claude-Code-Game-Studios26k—~1.4kAutomated safety check: PassMIT
RuView Onboarding Path Pickerruvnet/RuView97k—~333Automated safety check: PassMIT

Similar skills

  • Onboard

    alirezarezvani/claude-skills

    /cs:onboard — Founder interview that populates ~/.claude/company-context.md using the canonical 7-dimension cs-onboard schema.

    28k GitHub stars~1.3k tokensUpdated 1 mo ago
    Auto-check passed
  • Analyze an unfamiliar codebase and generate a structured onboarding guide with architecture map, key entry points, conventions, and a starter CLAUDE.md.

    277k GitHub starsUsed in 3 repos~2k tokens
    DevelopmentAuto-check passed
  • Onboarding

    sickn33/agentic-awesome-skills

    When the user wants to optimize post-signup onboarding, user activation, first-run experience, or time-to-value.

    47k GitHub starsUsed in 1 repo~1.8k tokens
    Auto-check passed
  • Contributor Onboarding Doc

    Donchitos/Claude-Code-Game-Studios

    Writes an onboarding document for a new contributor or agent, covering project state, conventions and priorities relevant to a chosen role.

    26k GitHub stars~1.4k tokensUpdated 3 days ago
    DevelopmentAuto-check passed
  • Zero-to-sensing path picker for RuView (WiFi-DensePose) — pick docker-demo, repo-build, or live-esp32 and run the next concrete step.

    97k GitHub stars~333 tokensUpdated today
    DevelopmentAuto-check passed
  • 分析一个陌生的代码库,并生成一个结构化的入门指南,包括架构图、关键入口点、规范和一个起始的CLAUDE.md文件。适用于加入新项目或首次在代码仓库中设置Claude Code时。

    277k GitHub starsUsed in 1 repo~1.1k tokens
    DevelopmentAuto-check passed

More from pandazki/pneuma-skills

All 30 skills in this repo
  • Pneuma Bansho

    pandazki/pneuma-skills

    Explain something by writing it on a board. An agent skill from pandazki/pneuma-skills.

    161 GitHub stars~6.9k tokensUpdated 2 days ago
    Auto-check passed
  • Pneuma Clipcraft

    pandazki/pneuma-skills

    AI-orchestrated video production on @pneuma-craft. An agent skill from pandazki/pneuma-skills.

    161 GitHub stars~7.5k tokensUpdated 2 days ago
    Auto-check: notes
  • Pneuma Lucid

    pandazki/pneuma-skills

    Pneuma Lucid Mode workspace guidelines. An agent skill from pandazki/pneuma-skills.

    161 GitHub stars~4.6k tokensUpdated 2 days ago
    Auto-check: warnings
  • Pneuma Plotwise

    pandazki/pneuma-skills

    Pneuma Plotwise workspace guidelines. An agent skill from pandazki/pneuma-skills.

    161 GitHub stars~8.9k tokensUpdated 2 days ago
    Auto-check passed
  • Pneuma Sprite

    pandazki/pneuma-skills

    Pneuma Sprite Mode workspace guidelines. An agent skill from pandazki/pneuma-skills.

    161 GitHub stars~16k tokensUpdated 2 days ago
    Auto-check passed
  • Pneuma Webcraft

    pandazki/pneuma-skills

    Pneuma WebCraft Mode workspace guidelines with Impeccable.style design intelligence.

    161 GitHub stars~7.5k tokensUpdated 2 days ago
    Auto-check: notes

Questions about Pneuma Project Onboard

What does Pneuma Project Onboard do?

Onboard an existing folder as a Pneuma project. An agent skill from pandazki/pneuma-skills. Pneuma Project Onboard is an agent skill from pandazki/pneuma-skills. Onboard an existing folder as a Pneuma project.

How do I install Pneuma Project Onboard in Claude Code?

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

How do I install Pneuma Project Onboard in Codex?

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

Can I use Pneuma Project Onboard 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 pandazki/pneuma-skills --skill pneuma-project-onboard -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/pneuma-project-onboard, .gemini/skills/pneuma-project-onboard, .github/skills/pneuma-project-onboard and .opencode/skills/pneuma-project-onboard in your project.

What does Pneuma Project Onboard need to run?

Going by SKILL.md and its folder, Pneuma Project Onboard needs the command-line tools its instructions call (bun) and credentials named OPENROUTER_API_KEY, FAL_KEY and FAL_API_KEY. Our summary lists: A credential in OPENROUTER_API_KEY; A credential in FAL_KEY.

Does Pneuma Project Onboard access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Pneuma Project Onboard safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Pneuma Project Onboard use?

Pneuma Project Onboard is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Pneuma Project Onboard use?

About 5.2k tokens (SKILL.md is roughly 21k 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 Pneuma Project Onboard?

Skills that share tags, products or a category with Pneuma Project Onboard: Onboard (alirezarezvani/claude-skills, 28k stars), Codebase Onboarding (affaan-m/ECC, 277k stars), Onboarding (sickn33/agentic-awesome-skills, 47k stars) and Contributor Onboarding Doc (Donchitos/Claude-Code-Game-Studios, 26k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Pneuma Project Onboard?

pandazki (a GitHub user) maintains it in pandazki/pneuma-skills, which has 161 GitHub stars. The repository holds 30 skills in this directory. The repository was last updated on October 9, 2026.

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