DRAGONS / FLIGHT RECORDER

The flight recorder for your coding agents.

Your agent said "done." It wasn't. You found out three days later. Every run writes a replayable record: who the agent was, what it was authorized to do, what it did, whether it is still alive. Check it yourself.

ZERO-DEPENDENCY REPLAY Download kit
curl -fsSLO https://dragons.dev/receipt-specimen.json
curl -fsSLO https://dragons.dev/replay.mjs
node replay.mjs receipt-specimen.json

Run the same refusal and recovery checks locally with Node. No account, install, or package registry.

01 / THE RUN

A run leaves a record you can check.

Dragons follows a local coding-agent run from activation to replay. The record names the agent, its authority, its work, and its last heartbeat instead of asking you to trust a dashboard.

CAUGHT-LIE SCENARIO Your coding agent says "done." Replay checks the run plan, work item, evidence, and last heartbeat against checked-in test records. If the record does not replay, the claim is not accepted.

  1. 01

    Dock

    Connect a local Claude Code or Codex session.

  2. 02

    Activate

    Identity and the manifest enter the record.

  3. 03

    Work

    Work as normal inside the declared boundary.

  4. 04

    Watch

    The last heartbeat shows alive versus silently dead.

  5. 05

    Replay

    Check the record against checked-in test records.

02 / EVIDENCE SURFACE

The first win is catching the lie.

A flight recorder earns its keep when it catches a false "done" or finds an overnight run that died before morning.

FALSE DONE

Replay the claim.

The record points from manifest to work item to evidence, so "done" can be checked against the run.

DEAD RUN

Find the silent stop.

The last heartbeat shows whether an overnight run is alive or needs recovery before morning.

OPERATOR BOUNDARY

Keep the decision yours.

Authority, evidence, and recovery stay visible to the operator instead of disappearing behind a vendor dashboard.

03 / THE SMALLEST USEFUL LOOP

One local dock. One replayable record.

run:
  target: Claude Code or Codex
  connect: MCP config line | SDK agent()
record:
  fields: run plan | instance | work item | heartbeat
replay:
  against: checked-in repo fixtures
  answer: alive | dead | done

A SMALLER PROMISE

Dock one local session. Replay one record. The proof path belongs in the repo and can be checked on your machine.

See the local path

04 / FAILURE RECORD

See the failure path before you trust the claim.

This is the public fixture the evaluator replays: one false "done" and one dead overnight run, with the decision each should trigger.

RECEIPT / TEST RECORD STRUCTURE ONLY
agent
coding-agent / identified
manifest
hash + one work item
authority
one repository task / granted
liveness
alive / heartbeat redacted
replay
agent → work → evidence → outcome
FALSE DONEREJECT / HOLD ACCEPTANCE

A completion claim with no matching artifact is refused, not accepted.

DEAD OVERNIGHT RUNRECOVER / BEFORE MORNING

heartbeat = stale

{
   "test_record": true,
  "production": false,
  "agent": "agent_019c8c0f2e2d7c1a8b6e2d4f9a0b1c3e",
  "manifest": "hash + one work item",
  "liveness": "alive / heartbeat redacted",
  "replay": "agent → work → evidence → outcome",
  "private_fields": "redacted"
}

05 / LOCAL DOCK

Work where your agent already works.

The product is a local SDK and MCP dock for coding-agent sessions. Connect Claude Code or Codex, work normally, then replay the record against checked-in test records.

01 / CONNECT

One local seam.

Use one local MCP config line or one SDK agent() call.

02 / WORK

Keep the run visible.

Each activation records the run plan, which agent ran, the work item, and the last heartbeat.

03 / CHECK

Replay the result.

When the agent says "done," check the record. It either replays or it does not.

import { agent } from '@dragons/sdk';

const run = agent({
  mode: 'persistent',
  venue: 'cloudflare',
  template: 'pi-mono',
  version: 'v1',
  dock: { org: 'mogos', project: 'fleet-of-one', role: 'worker' },
});

PUBLIC PATH / RUN IT NOW

The local SDK/MCP dock is the next product surface. Use the public replay kit above today; the local package and hosted endpoint are not part of this release.

Run what is public now

Need private local access? Email the operator. This is a human contact route, not an automated intake.

AVAILABLE NOW / FAILURE REPLAY You can run the deterministic false-done and dead-run scenarios today. They prove the delivered contract path, not a customer run.

06 / TRUTH FENCE

What is real in this release.

Static public surface plus a downloadable, deterministic false-done and dead-run replay. This release has no published SDK/package, installer, hosted endpoint, external users, customer outcomes, pricing, open-source release, or production-readiness claim. Next product evidence: one redacted replay from customer-owned or auditor-owned data.

STATUS / CONTRACT REPLAYEVIDENCE STAGED IN PUBLIC