Repository-native guidance for coding-agent work.

Turn agent work into a bounded repository contract with named validation and a written implementation summary.

A concise checklist for handing an implementation task to a coding agent and receiving work back with usable evidence.

Use coding assistants for implementation; keep the durable implementation contract and evidence in versioned repository files.

Keep planning and coordination in your tracker; govern the implementation contract and validation evidence in the repository.

Put executable validation in the task contract, require the observed result in the implementation summary, and treat missing evidence as incomplete work.

A Human-Agent-Contract is a chain of versioned repository artifacts from intent through validation evidence and acceptance.

Start with one bounded, reversible task and establish repository-visible validation and review evidence before expanding use.

A good task definition states the objective, implementation boundary, acceptance criteria, and executable validation before the agent changes code.

Review the declared boundary, changed paths, validation evidence, and risk areas first, then inspect code where the evidence or scope requires it.

Store durable intent, scope, validation, decisions, and implementation evidence in repository artifacts instead of relying on remembered chat context.

Escalate when scope, product intent, risk, or evidence can no longer be resolved within the task's declared boundary.

Use narrow tasks, least-privilege access, traceable evidence, and the same security and approval controls that apply to any other code change.

Require a written implementation summary with changed paths, validation results, unresolved risks, and an acceptance or follow-up decision.

Trace the specification through planning, then create tasks that have a single verifiable outcome, bounded paths, and explicit validation.

Coordinate agents through explicit task boundaries, dependency order, isolated working areas, and shared repository evidence.

Stop the task, preserve the evidence, compare the result to the declared boundary, and create an explicit corrective task or revert through normal review controls.

Make structure, scope checks, and validation deterministic; use models for interpretation, implementation, and human-supervised judgment where ambiguity remains.

No. A coding-agent workflow adds repository-level task and evidence structure; pull requests, review, and CI keep their existing roles.

Measure completed, accepted outcomes and their review and rework cost, not prompts sent or lines generated.

Declare allowed target paths before implementation, review changed paths against that boundary, and require escalation before the agent broadens scope.

A coding assistant performs work in a session; an agent workflow defines the durable repository contract, evidence, and review process around that work.

Plain-language explanation of the Human-Agent-Contract: explicit intent, scope, validation, evidence, and reconciliation for coding-agent work.

Why repository artifacts outlast chat transcripts for coding-agent planning, review, and handoff.

How Day Shift works alongside Jira and Linear instead of replacing them, managing the implementation contract inside the repository.

Replay a fresh Day Shift lifecycle, inspect generated artifacts, and play the implemented Tic-Tac-Toe result.

Frequently asked questions about Day Shift's repository-native workflow role.