Skip to content
Agentmart
Submit a tool

orchestrator

Use Snipara Orchestrator for explicit controller plans, task coordination, execution receipts and proof gates.

Skill by Snipara

IndexedMIT
Write a review

Chalk Labs

Need help deploying orchestrator?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/snipara-orchestrator

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use Snipara Orchestrator for explicit controller plans, task coordination, execution receipts and proof gates. Invoke only when the user requests orchestration or an authorized task requires it.

Use cases: productivity

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md and code. Checked Nov 1, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/orchestrator

What's inside

Skill name
orchestrator
Description
Use Snipara Orchestrator for explicit controller plans, task coordination, execution receipts and proof gates. Invoke only when the user requests orchestration or an authorized task requires it.
License
MIT
Instructions
63 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

Read `../../config.json` and require `modules.orchestrator: true`. Resolve the
plugin's `../../scripts/snipara.mjs` to an absolute path. Use
`node <script> orchestrator <arguments>` so the module gate and pinned runtime
are enforced. If absent, show `node <script> install orchestrator`; install only
when requested. Enabling this module never starts a controller or worker.

Before hosted or execution work in account mode with Snipara on, call
`snipara_stats` on this plugin's MCP connection and compare its `project_id`
to the explicit local profile `project`. Stop on missing configuration,
authentication failure or mismatch; use settings to align the profile with the
project selected in OAuth. Never take the project from another global MCP or
an inherited environment. Local status/help remains available without this
hosted check. With Snipara off, follow the restricted local-only behavior below.

1. Inspect `status --json` and the requested command's `--help` before work.
2. Discover native hosts with `host discover --adapter all`; discovery is not
   proof that a host can execute or that it is authorized for this task.
3. Translate an explicit goal into a controller plan, acceptance conditions and
   bounded budget. Start with observe autonomy unless execution is authorized.
4. Preserve stored autonomy ceilings, proof floors, permissions and evidence.
   Reconcile/preview before an explicit `--execute` action. Do not infer release,
   deployment, publishing or worker-messaging authority from plugin activation.
5. Capture execution receipts and review/proof evidence. Treat native execution,
   independent acceptance and live customer-path proof as separate outcomes.

Hosted htask operations require account/project credentials and the hosted
service. Plugin OAuth does not provision CLI credentials. With Snipara off, do
not invoke hosted queues or context; this version permits `status --json`,
help and version only.
Keep htask coordination here, not on Companion. Report unavailable capabilities
or missing proof without weakening an existing gate. Never fabricate credentials,
approval, acceptance, runtime continuation or billing measurements.

Before repository tests or executable validation, use the module's `validation gate`
with a Sandbox request and explicit workflow/lot scope. Probe required capabilities,
not merely installation. If Sandbox is absent/inaccessible/incompatible, stop that
validation and ask the user: another test environment, install/configure Sandbox,
or continue without Sandbox (clarify local tests versus no tests). No silent host
fallback and no automatic installation. Record the explicit choice with
`validation choose --acknowledge-user-choice`, only after the user actually chose.
Keep that choice scoped to the workspace, lot, profile and permissions; preserve
it across resume without resetting budgets. A missing response is not approval.
Use `validation run` and `validation verify` for exact-snapshot Docker evidence.
Keep proof floors and independent review requirements even when skipping tests.
Static reads/documentation do not require a container. Sandbox is standalone;
no account or inference is required for its deterministic validation commands.

For new executable Controller lots using Sandbox, install a policy-owned proof
floor using `--required-proof "snipara-sandbox validation verify --request ...
--run-id ... --store ..."` once the run is available and the exact request is
frozen. A verifier failure blocks integration through the existing proof gate.
Never count `validation gate`, `choose`, `diagnose`, or a saved choice as a passed
validation. Existing Controllers keep their stored proof floors; do not silently
rewrite their policy. The generic legacy local proof runner remains available
for explicitly authorized alternative environments, not an implicit fallback.

The launcher excludes environment credentials by default. For a hosted command
that needs a Snipara key, explain sharing with Orchestrator and obtain explicit
user authorization before placing `--with-snipara-credentials` immediately after
`orchestrator`. Permission to orchestrate alone does not authorize credential
sharing. Never ask for or print keys or forward unrelated provider credentials.

From Snipara/snipara-claude, MIT.

More from Snipara/snipara-claude

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Oct 24, 2026 (8 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

Same skill elsewhere

Copies of this SKILL.md published by others; search shows only one of them.

More from Snipara

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get orchestrator running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying orchestrator?

CallGet help
Cite this page

orchestrator skill. Agent Mart. https://agentmart.dev/skills/snipara-orchestrator (last updated 2026-10-24)