Save a completed task as a reusable prompt

Choose the latest task in this conversation with user confirmation or actual verification evidence. Identify the task and evidence briefly. Exclude abandoned attempts and unverified conclusions. If no task qualifies, stop without requesting more history.

Extract its objective, necessary inputs, constraints, output format and completion criteria. Remove incidental names and dates while preserving meaningful conditions. Have the future agent infer project context where possible; clearly identify inputs that really must be supplied.

Deliver the complete reusable prompt, not an outline. Do not create a Skill by default; mention suitability only if the record establishes a stable recurring workflow.

Check the prompt against this task's requirements, state its limits, and remove irrelevant personal information and credentials.

Execution contract:
Start directly. Do not ask me to approve a plan, choose files or fill in a template. Use only this project, this conversation and materials already provided; do not search unrelated projects or private history. Complete what the evidence supports and report missing inputs. If the task is inapplicable, explain why and stop without inventing a result.
Preserve existing files and uncommitted changes. Include complete proposed content in your reply; standalone artifacts may be saved under a new, unused filename. Do not install dependencies, change the environment, commit, push or deploy. Respect access and permission limits and report blocked work. Check your result once, deliver it and stop without asking whether to continue.