ClawKeeper
ClawKeeper checks proposed Claude agent actions against explicit rules before a model call. Autonomous spend defaults to zero, and the database enforces an append-only audit log.
Project details
- Classification
- Open-source project
- Repository
- Alexi5000/ClawKeeper
- Language
- TypeScript
- License
- MIT
- Status
- Active development
- Tests
- 3 test notes
- Last verified
- 2026-08-06
- Related write-up
- alexcinovoj.com
What it is
I built ClawKeeper to check proposed actions before Claude agents make model calls. An operator must explicitly raise the autonomous budget from zero. The database enforces an append-only audit log. You can inspect the policy engine, workflow files, and reference finance implementations in the repository.
Architecture
- Policy engine evaluates a proposed tool call against deterministic rules before any provider request is issued.
- Agent clusters are grouped by responsibility: intake, validation, reconciliation, reporting, compliance.
- Postgres with row-level security carries tenant isolation; the audit table is append-only and enforced in the database.
- Bun and Hono on the service side, React on the dashboard side, Docker Compose for local bring-up.
- Skills and steering documents (SKILLS.md, STEERING.md, AGENTS.md) are versioned so readers can inspect changes to agent instructions.
operator request
|
v
+-----------------+ deny
| policy engine |----------------> audit log (append-only)
| (deterministic) |
+--------+--------+
| allow
v
+-----------------+ +----------------------+
| budget ceiling |---->| provider call |
| (default $0) | | (Claude) |
+-----------------+ +----------+-----------+
|
v
audit log (append-only)Tests and gates
- A test suite runs in the repository's GitHub Actions workflow (.github/workflows/ci.yml); see the latest run for current results.
- Lint and type checks are wired into the same workflow via eslint.config.js and the package scripts.
- CODEOWNERS, CONTRIBUTING.md, and SECURITY.md are present, so review and disclosure paths are defined.
Limits
- You need to deploy and operate this open-source project yourself. No managed deployment is provided.
- The policy engine is deterministic by design: it does not reason about intent, so policies must be written explicitly.
- The finance workflows are reference implementations. They are not accounting certifications or independently validated accounting logic.
Source files and examples
Related methods
Acceptance tests for agent work →
Executable checks test the task and detect regressions.
Evaluation tests for releases →
A fixed evaluation suite checks release requirements.
MCP tool permissions →
The server checks which tools a caller may use.
Recording agent runs →
A shared trace ID connects the records of a run.
Spending limits for agent runs →
Budget checks run before provider requests.
Pausing for human approval →
The run waits for approval before the action executes.
Related projects
TechTide Harness Kit →
TechTide Harness Kit uses JSON Schema to declare skill permissions, risk tiers, network access, and approval requirements. Automated checks validate declarations and run prompt-injection fixtures.
CipherClaw →
CipherClaw analyzes recorded agent activity without model calls. It builds causal graphs, scores possible causes, and flags failure patterns.
Swarm 357 →
357 Claude agent roles across six business layers, with portable memory, enforced cost caps, a bash policy gate, and durable checkpoints.
FintheFinder →
FintheFinder pauses research runs for human approval and audits citations deterministically. Its repository includes contracts and test files.
I document the project here. TechTide AI publishes current service scope, prices, and terms.
Systems Decision Audit at TechTide AI→© 2026 Alex Cinovoj · TechTide AI · Columbus, OH