Skip to main content

Deliver Promised Work

Complete one promised job to its frozen acceptance condition. Completion is not acceptance.

Use when. A promise, owner, inputs, due condition, and acceptance test are known.

Do not use when. The promise is still ambiguous, or you want the model to redefine scope, spend, or the consumer.

Your input. Promise and version; beneficiary; acceptance condition and due date; inputs and dependencies; permitted actions; escalation path.

Your quality check. The result can be checked against the frozen condition. Deviations are visible. Consumer acceptance stays blank.

Put this to work

Coordinate one promise to acceptance

For the accountable delivery owner

Copy this prompt. Paste it with the relevant source material into Claude, ChatGPT, or any AI assistant, then review the result before acting.

# Role and job
Help an accountable delivery owner coordinate one promised job to its frozen acceptance condition.

# Outcome
Produce a Delivery Run Brief and acceptance handoff. Success means the named consumer can accept or reject the result against the original condition, with deviations visible.

# Inputs
- Promise and version: [paste]
- Beneficiary and consumer: [name]
- Acceptance condition and due date: [paste]
- Inputs, dependencies, and responsible producer: [paste]
- Permitted actions and constraints: [paste]
- Escalation path and evidence destination: [paste]

# Boundaries
You may coordinate, draft, check, or perform only explicitly delegated steps. Do not redefine the promise, acceptance, priority, spend, risk, or consumer.

# Output
Return: frozen promise; execution checklist; dependency state; deviations; result; acceptance evidence; consumer decision placeholder; follow-up owner; and review date.

# Stop and recovery
Stop for a blocked critical input, unsafe condition, scope conflict, unapproved change, or uncheckable result. Preserve work, name the blocker, and return the smallest recoverable next step.

# Final check
Verify the result against the frozen acceptance condition and leave acceptance to the named consumer.

Human decision — cannot be delegated. The named consumer accepts, rejects with evidence, or requests an authorised change. No clinic CTA.

Candidate teaching projection; contextual recipe unproved. The JTBD Ledger keeps this route as a gap until a contextual owner accepts the method and retains decision evidence.

Outcome

The named consumer receives work that meets the agreed acceptance condition, with deviations visible.

Use When

Use when a promise, owner, inputs, due condition, and acceptance test are known.

Inputs

Promise and version, beneficiary, acceptance condition, deadline, inputs, producer, dependencies, escalation, and evidence capture point.

Recipe

Copy the prompt. Restate the promise. Execute the smallest complete job. Leave consumer acceptance blank.

Human and AI Control

An assistant may coordinate, draft, check, or perform explicitly delegated steps. It may not redefine the promise, acceptance, priority, spend, risk, or consumer. Stop for a blocked critical input, unsafe condition, scope conflict, or an uncheckable result.

Output and Handoff

Accepted work moves to compare outcome and correct method. Rejected work returns to the delivery owner.

Evidence and Stop

Stop on a protected-constraint breach or when acceptance is unavailable. Completion is not evidence of acceptance or useful consequence.

Reliability

Skill and tool: promise-to-acceptance delivery using only authorised coordination and checking tools. The run brief is the pattern; frozen acceptance and consumer review are the protocol. Retain deviations and the consumer decision as evidence. Accept only the consumer's recorded decision.

Next

Correct one failed step or acceptance definition before repeating. Return promise ambiguity to the commercial owner.