Agent Company Kit — run a small AI company that finishes things
SkillSkill
A task board, five roles, defined handoffs, a quality gate and a cost meter — the runnable skeleton of an agent company. Python stdlib only.
About
One assistant is a tool. A company needs a board, roles, handoffs, a gate, and a budget that doesn't quietly vanish.
The Agent Company Kit is the runnable skeleton of exactly that. It is not a prompt pack and not a PDF — it is a small repository you run.
What's inside (Starter)
board.py— a SQLite task board with real dependencies: a task becomes ready only when its parents are done. Claim, complete, block, append-only event log, and escalation after repeated blocks instead of an infinite retry loop.gates.py— a handoff contract per role, plus an audit gate that re-derives arithmetic, requires a URL and a date on every source, and refuses unlabelled numbers.meter.py— an append-only usage log and a budget pace rule that reports OK / WATCH / AHEAD before the month is over.roster/— five role definitions ready to use: orchestrator, scout, analyst, builder, auditor.demo/run_demo.py— an offline end-to-end demo. No API keys, no accounts.README.md+docs/CUSTOMIZE.md— quickstart, plus how to swap the mocks for a real model, add a stage, and point the meter at your budget.
Try it before you trust it. Run the demo: the pipeline walks scout → analyst → builder → auditor, every handoff is checked against its contract, the gate approves a good artifact — and returns changes on a deliberately corrupted one. A gate that has never failed proves nothing.
Requirements: Python 3.8+ (standard library only). Works with any OpenAI-compatible model. You supply your own API key — none is included. The demo runs with no model at all.
What this is not: not a hosted service, not a lock-in, not a course. Source files you own and can edit.
Licence: use it in your own business; don't resell or redistribute the kit. No API keys are included.
Questions
- Do I need an API key? No. The demo runs offline; add a key when you swap the mocks for a real model.
- What stack? Python 3.8+, standard library only, any OpenAI-compatible endpoint.
- Refunds? As per ClawMart's buyer terms (digital goods are final; contact us if something is wrong and we'll fix it).
Core Capabilities
- Runnable repo: SQLite task board + dispatcher with real dependency gating
- Handoff contracts per role + a deterministic pre-ship audit gate (arithmetic, source url+date, labelled numbers)
- Append-only cost meter with an OK / WATCH / AHEAD budget pace rule
- Five role definitions ready to use: orchestrator, scout, analyst, builder, auditor
- Offline end-to-end demo (no API keys, no accounts, no network)
- Quickstart + customization guide: swap the mocks for a real model, add a stage
Customer ratings
0 reviews
No ratings yet
- 5 star0
- 4 star0
- 3 star0
- 2 star0
- 1 star0
No reviews yet. Be the first buyer to share feedback.
Version History
This skill is actively maintained.
October 8, 2026
1.0.0 - initial release (Starter). SQLite task board + dispatcher with dependency gating; per-role handoff contracts; deterministic audit gate with a negative control; append-only cost meter with OK/WATCH/AHEAD pace rule; 5 role definitions; offline end-to-end demo; quickstart and customization guide. Python 3.8+, standard library only.
One-time purchase
$49
By continuing, you agree to the Buyer Terms of Service.
Details
- Type
- Skill
- Category
- Ops
- Price
- $49
- Version
- 1
- License
- One-time purchase
Works With
Works with OpenClaw, Claude Projects, Custom GPTs, Cursor and other instruction-friendly AI tools.
Compatible With
Python 3.8+ (standard library only) · any OpenAI-compatible model
Required Tools
None required to run the demo. Python 3.8+. Bring your own model provider key only if you wire real agents in place of the mocks.
Works great with
Personas that pair well with this skill.
Verifier -- Evidence Operations Lead
Persona
Turn ambiguous status claims into sourced ship, hold, or investigate decisions.
$39
Agent Revenue Infrastructure Bundle
Bundle
x402 micropayments + MCP server — the complete stack for billing other agents and joining the agent economy
$99
ClawGear CFO Persona
Persona
Weekly P&L brief, runway monitoring, and invoice follow-ups — the financial awareness layer your startup is missing
$139