Skip to content
Skills Library
Programming · Review & audit

treeship-dev

/treeship-dev

Work on the Treeship source repository safely. Use when Codex or another coding agent is asked to modify, review, document, or test the zerkerlabs/treeship repo, especially packages/core, packages/cli, packages/hub, docs, SDKs, MCP bridge, DSSE artifacts, approvals, handoffs, Hub/Dock/DPoP, or cryptographic verification behavior. This skill provides repo scope, required read order, crypto invariants, CLI UX rules, and validation commands for contributors.

Repository
treeship ↗
Licence
Apache-2.0
Last repository change
27.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 27.10.2026, without references to retired models or APIs.

Licence

Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.

What it does and when to use it

Work on the Treeship source repository safely. Use when Codex or another coding agent is asked to modify, review, document, or test the zerkerlabs/treeship repo, especially packages/core, packages/cli, packages/hub, docs, SDKs, MCP bridge, DSSE artifacts, approvals, handoffs, Hub/Dock/DPoP, or cryptographic verification behavior. This skill provides repo scope, required read order, crypto invariants, CLI UX rules, and validation commands for contributors.

Original author description

Work on the Treeship source repository safely. Use when Codex or another coding agent is asked to modify, review, document, or test the zerkerlabs/treeship repo, especially packages/core, packages/cli, packages/hub, docs, SDKs, MCP bridge, DSSE artifacts, approvals, handoffs, Hub/Dock/DPoP, or cryptographic verification behavior. This skill provides repo scope, required read order, crypto invariants, CLI UX rules, and validation commands for contributors.

How it works

  1. Start Here
  2. What Treeship Is
  3. Non-Negotiable Invariants
  4. Repo Map
  5. Preferred Read Order
  6. CLI UX Rules
  7. Validation
  8. Working Style
Original SKILL.md headings · 107 lines, 534 words
SKILL.md preview
---
name: treeship-dev
description: Work on the Treeship source repository safely. Use when Codex or another coding agent is asked to modify, review, document, or test the zerkerlabs/treeship repo, especially packages/core, packages/cli, packages/hub, docs, SDKs, MCP bridge, DSSE artifacts, approvals, handoffs, Hub/Dock/DPoP, or cryptographic verification behavior. This skill provides repo scope, required read order, crypto invariants, CLI UX rules, and validation commands for contributors.
---

# Treeship Dev

## Start Here

Use the local `zerkerlabs/treeship` checkout as the workspace root. Stay scoped to that repo unless the user explicitly asks to inspect sibling repos.

Read these files before code changes:

1. `AGENTS.md`
2. `ONBOARDING.md`
3. `docs/quality/ai-assisted-development.md`

Treat `AGENTS.md` as the source of truth. If these skill instructions conflict with `AGENTS.md`, follow `AGENTS.md`. The code-quality policy in `docs/quality/ai-assisted-development.md` is not optional; it targets the specific failure modes AI-generated code is prone to in this cryptographic codebase.

## What Treeship Is

Treeship is a portable trust layer for AI agent workflows. Actions, approvals, handoffs, receipts, bundles, and related records are signed artifacts that must verify without trusting hosted infrastructure.

The core loop is:

```bash
treeship wrap -- your-command
treeship hub push <artifact-id>
```

## Non-Negotiable Invariants

Preserve these exactly:

- DSSE PAE format: `"DSSEv1" SP LEN(payloadType) SP payloadType SP LEN(payload) SP payload`
- Artifact ID derivation: `artifact_id = "art_" + hex(sha256(PAE_bytes)[..16])`
- Statement structs do not contain an `id` field. IDs live on records/sign results.
- Envelope JSON uses DSSE camelCase: `payload`, `payloadType`, `signatures`.
- Approval nonce binding is enforced: `action.approvalNonce == approval.nonce`.
- Hub/Dock auth uses DPoP with a separate dock keypair. Do not add session tokens.
- Rekor anchoring is best-effort and must not make artifact push fail when Rekor is down.
- TypeScript SDK verification must use the WASM verifier path, not a `treeship` subprocess.

## Repo Map

- `packages/core`: Rust core library, attestation, statements, keys, storage, verifier, Merkle.
- `packages/cli`: Rust CLI, command tree, wrap/status/verify/hub/ui/otel commands.
- `packages/hub`: Go Hub server and SQLite-backed API.
- `packages/core-wasm`: Browser verifier.
- `packages/sdk-ts`: TypeScript SDK.
- `packages/sdk-python`: Python SDK.
- `bridges/mcp`: MCP bridge.
- `docs`: Fumadocs documentation.
- `skills`: portable agent skills.
- `plugins`: Codex plugin candidates.

## Preferred Read Order

For core protocol or CLI work, read:

1. `packages/core/src/attestation/pae.rs`
2. `packages/core/src/attestation/sign.rs`
3. `packages/core/src/attestation/verify.rs`
4. `packages/core/src/statements/mod.rs`
5. `packages/cli/src/main.rs`
6. `packages/cli/src/commands/wrap.rs`

For Hub work, read `packages/hub/main.go` and the relevant package under `packages/hub/internal`.

For docs-only work, still read `AGENTS.md`, then edit only the relevant pages under `docs`.

## CLI UX Rules

When changing CLI behavior:

- Keep commands noninteractive.
- Provide `--format json` where command output needs stable machine-readable output.
- Return useful exit codes: `0` ok, `1` error, `3` not initialized, `4` usage.
- Include a concrete fix in errors.
- Print a dim next-step hint on success.
- Respect `NO_COLOR` and `--no-color`.
- `treeship wrap` must propagate the subprocess exit code.

## Validation

Choose focused checks for the files changed:

```bash
cargo test -p treeship-core
cargo test -p treeship-cli
cargo test
```

For Hub:

```bash
cd packages/hub
go test ./...
go build ./...
```

For TypeScript or MCP packages, inspect package scripts first, then run the relevant npm test/build command from that package directory.

## Working Style

Keep edits narrow. Do not refactor unrelated crypto, storage, or schema code while solving a task. Avoid changing generated IDs, signed payload shape, schema field casing, or verification semantics unless the user explicitly asks and the change is reconciled with `AGENTS.md`.

The full rules for scope, silent fallbacks, verifier preconditions, wire-bound dispatch fields, TOCTOU, RNG, test rigor, and theater commits live in [`docs/quality/ai-assisted-development.md`](../../docs/quality/ai-assisted-development.md). Follow that policy.

Use direct, concise writing. Avoid corporate phrasing.

Versions

RepositoryScoreLicence
This page · zerkerlabs/treeship100Apache-2.0
zerkerlabs/treeship100Apache-2.0

Database snapshot: 13.11.2026 · package version 9f7c8909c800 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100treeship-report
    Use when the user asks for "the URL", "a shareable link", "the proof I can send to <someone>", or otherwise wants the current Treeship session published as a session report on treeship.dev.
  • 100treeship-session
    Use when the user finishes a meaningful unit of work (bug fix, feature, migration, refactor) and you need to close out the current Treeship session with a headline, summary, and review note that captures what actually happened.
  • 100repo-source-code-test-packages
    Write unit tests for Formisch packages (packages/core and packages/methods) with proper TypeScript types. Use when creating new tests, fixing type errors in tests, or adding test coverage for core/methods functions.
  • 90dev-hub
    Set up the project layout for the RayStack dev flow, with AGENTS.md, DECISIONS.md and the docs folders, in one repo or in a code-free hub across repos. Use for "set up the dev flow", "start a project hub", "add DECISIONS.md", or when dev finds no hub.
  • 90reference-core
    Explains the mental model and architecture of the code under `packages/core`. You MUST use this skill any time you plan to work with code in `packages/core`
  • 100review-core-sdk
    Use when reviewing a change, diff, or PR to the SODAX core SDK (packages/sdk plus the type and wallet packages it touches) — checking correctness, scope, architecture, types/safety, tests, source-derived config, secrets, and AI-doc faithfulness before it lands. Triggers on…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.