Be-Ambitious Harness

Three roles and one human. Planner turns the task into a prompt with two bars and a gate check. Worker executes the prompt, starting with the check. Verifier grades the result seeing only the goal, requirements, criteria and facts. Human chooses the scheme and cost limit — this is not delegated.

Each role reads only its own files. This is deliberate: instruction following breaks down past 5–6 simultaneous constraints, so structure comes from files and templates, not from rules held in mind.

Flow

0. Intake. Collect four fields: task, context, mode (keep scope / ambitious extension / recommend), stop_after_analysis (yes/no). Defaults if empty: recommend, no.

1. Routing. Open scenarios/_index.md, pick a scenario by signals. Torn between two — take the one where the result lives. None fits — generic.md.

2. Analysis and prompt. Read agents/planner.md and the scenario file. Run the analysis; scenario hints serve their matching points, not all at once. Output questions and the prompt per core/prompt-template.md. With stop_after_analysis = yes, stop after analysis and questions; continue only after the human answers.

3. Pre-flight check. Run the prompt through checklists/pre-flight.md. Any "no" — back to step 2.

4. Worker. New session or subagent. Pass: prompt, materials, the scenario's "For the worker" block, agents/worker.md, core/gate.md, core/report-format.md. Nothing from the analysis.

5. Verifier. Separate session, preferably another model. Pass: prompt sections 1, 2, 4, 6, result, materials, the worker's evidence log, agents/verifier.md, core/report-format.md, tools/gate-check.py. Sections 3, 5, 7, the worker's reasoning and the planner's analysis stay with you.

6. Acceptance. Run checklists/post-run.md. To the human: result, verifier's verdict, list of unverified items. The decision "accept / rework / stop" is theirs.

Harness rules

• Mandatory requirements of the original task pass through every stage verbatim.
• Acceptance criteria and the gate criterion are fixed at step 2 and only read afterwards.
• Work specific to the ambitious branch starts after the gate outcome is logged.
• The verifier does not fix. The worker does not grade itself in place of the verifier.
• Any stage lacking materials requests them from the human and waits — this is not a gate outcome.

Files

• agents/ — role instructions: planner.md, worker.md, verifier.md
• core/ — shared definitions: gate.md, prompt-template.md, report-format.md
• scenarios/ — domain hints: _index.md and six scenarios
• checklists/ — pre-flight.md, post-run.md
• evals/ — test tasks and scorecard for the harness itself; run after any edit