Data platform engineer — seven years building cloud-native data and ML platforms in Python, Spark, and Terraform, and lately the governance layer that lets LLM agents do real engineering work without anyone having to trust them blindly.

Agents pull the oars. You hold the tiller.

Docs · Install · Releases · PyPI

Coxswain builds software with a crew of AI agents and keeps a person on the tiller. You describe a change; it is filed as a work item, sized (one ticket, or an epic broken into phases and tasks), planned, built in its own worktree under a dollar budget, linted and run against the project's own tests before a reviewer spends a token, reviewed by two independent reviewers who are made to disagree, arbitrated, validated against measured evidence rather than the builder's word, and handed back as a pull request. A cartridge's gate says how far the harness may go on its own; the default is ticket: it proposes, a person merges. One session holds the leader lock and drives the landing loop; every run leaves a record you can read, drill into, and get a notification from.

uv tool install coxswain-tools     # or: brew install ppfenning/coxswain/cox
cox setup doctor
cox                                # the coxswain session: file, launch, land

Status: beta, 0.2.0. The loop runs daily on its own repositories, which is also how its defects get found — and fixed, by the loop, as pull requests it opens against itself.

Agents should earn autonomy the way engineers do — by track record, in writing, revocably. Each repository is usable alone; together they read as one sentence.

flowchart LR
    UMB["coxswain<br/>the umbrella<br/>docs · manifest · one-line install"]
    CART["coxswain-cartridges<br/>who a run works for<br/>roles → skills · tier → model · where writes land"]
    GRAPHS["coxswain-graphs<br/>what runs, and the harness that runs it<br/>sequence · the gate · the ledger · worktrees"]
    CREW["coxswain-crew<br/>who speaks<br/>named seats · write authority · voices"]
    TOOLS["coxswain-tools<br/>the cox command<br/>run records · landing · the leader lock · cox home"]
    UMB -- "pins every component" --> CART
    CART -- "roles, tiers" --> GRAPHS
    CREW -- "seats bound by a cartridge's cast: block" --> CART
    GRAPHS -- "run records, usage, ledger" --> TOOLS
    TOOLS -- "files, launches, lands" --> GRAPHS

coxswain · coxswain-cartridges · coxswain-graphs · coxswain-crew · coxswain-tools

• A graph owns sequence and writes nothing; one harness owns every consequence — the gate, the worktree, the checks, the append-only ledger. A new graph inherits all of it by existing rather than by remembering to.
• Autonomy is earned per kind of write against the ledger, scoped to one cartridge hash and one provider profile — and an auto-applied write records no ledger row, so nothing can ratchet up its own trust. Demotion comes from measured signals, never model opinion.
• Every change arrives as a pull request, rebased onto the target's default branch and held until the project's own checks pass. How far the harness may go on its own is a cartridge policy with four named levels, tighten-only through the chain, never a hard-coded behaviour.
• An epic driver takes an initiative from idea to a stack of pull requests — phases in dependency order, tasks fanned out in parallel worktrees, each under its own budget ceiling.
• Two runners, one contract. Nodes run against the Messages API, or as headless Claude Code sessions on a subscription with per-role tool grants, a disposable scratch tree for the builder, dollar ceilings per tier, and a turn-by-turn trace. Every run writes what each node cost.
• A standing cast of named seats — recon, triage, review, build, ops, the board, writing, the scribe — defined with no employer inside them, and bound to a team's skills by a cartridge. Voice is decoration: every seat works typed.
• Deterministic work is a tool, not a turn. Summing a run's cost, counting what a traced node did, cleaning up after a run, drawing the live table: each is a tested function a seat calls, so no tokens are spent producing the same answer twice.