Supplies an AI coding agent with the facts, rules, boundaries, and examples needed to apply agent verification rules consistently. It fixes adoption boundaries and observable failure evidence while keeping environment-specific limits configurable.
Package status: reference context ready for human review. The contract and test scenarios are complete, but no claim is made that an adopting implementation has passed them.
Decision
For Agent verification rules, adopt version-controlled repository instructions scoped by path and task. Implement the behavior through one authoritative write path, typed interfaces, explicit state, and idempotent side-effect handling. This package is a reference contract: the adopting team must choose numeric limits, providers, jurisdictions, and operational owners from evidence in its own environment.
Scope
- The Agent verification rules actors, inputs, outputs, states, policy or schema versions, and externally visible outcomes described by this implementation block.
- Primary and alternate interfaces, background work, caches, integrations, support paths, and evidence that can exercise the same boundary.
- Adoption-specific configuration, rollout, recovery, and verification responsibilities needed to apply the reference safely.
Outside this block
- Choosing a universal vendor, framework, jurisdiction, numeric threshold, retention period, or service-level objective for every adopter.
- Claiming that packaged scenarios have run against a downstream implementation or that this reference grants legal, security, or accessibility certification.
Contract
- Declare instruction precedence, applicable paths, required checks, prohibited changes, and the evidence expected in handoff.
- Derive commands, versions, architecture, and file locations from repository sources rather than memory or a neighboring project.
- Keep executable verification separate from prose claims and record when a check was not run or could not complete.
- Require the agent to preserve unrelated changes and to stop when credentials, destructive scope, or product authority is missing.
- Validate inputs once at the boundary and preserve the normalized representation used by downstream logic.
- Persist the business transition before or atomically with an outbox record; consumers deduplicate by durable operation identity.
- List mandatory commands by changed path and risk, expected success evidence, time or environment limits, and how to report not-run or inconclusive checks.
- Agents may not convert static package validation into a claim that target tests passed, and must surface failures with the exact command and relevant output.
Implementation guidance
- Ship additive storage and interface changes first, backfill observably, then retire old paths after compatibility evidence.
- Model Agent verification rules inputs, outputs, actors, states, invariants, side effects, and evidence before selecting framework or vendor details.
- Store the applicable Agent verification rules contract or policy version with material state so migrations, replay, and support decisions remain attributable.
- Introduce the path behind controlled rollout, compare expected and observed outcomes, and keep a tested recovery route until adoption evidence is complete.
Failure handling and safeguards
- Conflicting or stale context is surfaced as a blocker with exact sources; the agent does not invent a tie-breaker that changes project behavior.
- Invalid or contradictory Agent verification rules state is rejected or quarantined; the implementation never guesses a value that broadens authority or duplicates an effect.
- Retries are bounded and use durable identity; after exhaustion, work reaches an inspectable terminal state with an accountable owner.
Verification and operations
- Inject dependency failure before and after persistence and prove that retry converges on one externally visible outcome.
- Monitor Agent verification rules success, denial, validation failure, dependency failure, retry exhaustion, and recovery by contract version without sensitive payload dimensions.
- Re-run paired positive and negative fixtures after policy, schema, dependency, migration, or boundary changes and preserve the resulting evidence.
Adoption assumptions
- Names and values in the example are a concrete fixture, not universal defaults; adopters replace them through documented evidence and ownership.
- The adopting system can provide authenticated identity, durable operation or record versions, bounded telemetry, and a controlled path for change.
The executable-looking examples in this package are fixtures and acceptance contracts. Run the collection validator to check structure and metadata, then translate and execute the scenarios in the target repository before recording implementation evidence.