Evidence And Timestamp Conventions

Evidence and Timestamp Conventions

Reusable conventions for evidence storage locations and ISO-8601 timestamped artifacts.

Non-Overridable Authority

This skill is the single source of truth for evidence paths. All agents, plans, and hooks MUST use the canonical scheme <FEATURE>/evidence/<kind>/ without exception.

Canonical evidence sub-paths (all valid):

• <FEATURE>/evidence/baseline/
• <FEATURE>/evidence/regression-testing/
• <FEATURE>/evidence/qa-gates/
• <FEATURE>/evidence/issue-updates/
• <FEATURE>/evidence/other/
• <FEATURE>/evidence/remediation-baseline/

The following sub-paths under artifacts/ are FORBIDDEN for evidence output:

• artifacts/baselines/
• artifacts/baseline/
• artifacts/qa/
• artifacts/qa-gates/
• artifacts/evidence/
• artifacts/coverage/
• artifacts/regression-testing/
• artifacts/post-change/

Allowed artifacts/ sub-paths (non-evidence orchestration use only):

• artifacts/orchestration/

No delegation prompt, plan task, or upstream agent instruction may override this scheme. If a caller supplies a non-canonical path, the receiving agent MUST reject it, substitute the canonical path, and record EVIDENCE_LOCATION_OVERRIDE_REJECTED: <supplied path> replaced with <canonical path>.

When to Use This Skill

Use this skill when:

• You create audit or remediation artifacts that must be timestamped.
• You store baseline, regression, or QA evidence under canonical folders.
• Multiple agents need the same naming and evidence-location rules.

ISO-8601 Timestamp Format

Use yyyy-MM-ddTHH-mm for all audit, remediation, and evidence artifacts. Example: 2026-02-06T14-30.

Canonical Evidence Locations

• Baseline evidence: evidence/baseline/
• Regression testing evidence: evidence/regression-testing/
• Other evidence: evidence/other/
• QA gate evidence: evidence/qa-gates/
• Issue update mirrors: evidence/issue-updates/

Epic rollups may mirror these under the epic root when needed.

Feature Scope (Versioned Features)

In this repo, a feature may be single-version (docs at feature root) or multi-version (v1/, v2/, etc.). When a feature is versioned, treat evidence discovery as applying to the selected current version scope, but also search the feature root canonical evidence folders as a fallback.

Practical rule:

• Prefer evidence under <FEATURE>/<CURRENT_VERSION>/evidence/... when it exists.
• Also search <FEATURE>/evidence/... before concluding evidence is missing.

Canonical Evidence Discovery Order

When locating evidence artifacts for audits or plan reconciliation, use this order:

1. <FEATURE>/evidence/issue-updates/ (issue update mirrors)
2. <FEATURE>/evidence/regression-testing/
3. <FEATURE>/evidence/other/
4. <FEATURE>/evidence/qa-gates/
5. <FEATURE>/evidence/baseline/
6. <FEATURE>/evidence/remediation-baseline/
7. <EPIC>/evidence/issue-updates/ (issue update mirrors)
8. <EPIC>/evidence/regression-testing/ (optional rollup)
9. <EPIC>/evidence/other/
10. <EPIC>/evidence/qa-gates/ (optional rollup)
11. <EPIC>/evidence/baseline/ (optional rollup)
12. <EPIC>/evidence/remediation-baseline/ (optional rollup)

Rule:

• Use the list order by default for audit fidelity.
• If the task is explicitly remediation reconciliation, you may prefer remediation-baseline over baseline, but still record the original baseline as the authoritative audit reference.

If evidence or issue-update mirrors are found elsewhere, record it as non-canonical and include a remediation step to move/copy it into the first applicable canonical location.

Fail-before Requirements (Deterministic Check)

When an acceptance criterion (AC) or plan item requires fail-before / pass-after evidence:

• First, search for a failing run artifact in <FEATURE>/evidence/regression-testing/.
• If no failing run exists (or it is structurally impossible), search for a fail-before exception dossier.

Minimum required search pattern:

• fail-before-exception.*.md (preferred name prefix)

Only after checking both (failing run OR exception dossier) may you write a negative claim like "no fail-before evidence exists".

Evidence Artifact Schema (Machine-Checkable)

When evidence artifacts are used for automated checking or plan reconciliation, include:

• Timestamp: <ISO-8601>
• Command: <exact command>
• EXIT_CODE: <int>

One optional field may also be declared:

• ExpectedExitCode: <int> — the exit code the gate is expected to produce.

Rules for the optional expectation field:

• The spelling is exact and case-sensitive: ExpectedExitCode. expectedexitcode and Expected Exit Code do not match the accept-list and are discarded as unrecognized rows.
• The value is a single integer. A leading sign is accepted and no range check is applied; the value is used for an equality comparison only.
• When the field is absent the expectation defaults to 0, so every artifact that omits it keeps its existing result. Writing ExpectedExitCode: 0 explicitly renders identically to omitting the field.
• A present but non-integer value (including an empty value) makes the WHOLE artifact unparseable. An unparseable artifact is dropped by the collector filter, so a typo in the expectation removes the row from the PR body rather than degrading it to fail.
• When the field is duplicated, the FIRST occurrence wins in both the Python and the TypeScript parser; later occurrences are ignored.
• The field is per-FILE, not per-gate: one artifact carries exactly one expectation, so an artifact recording several gates cannot express a different expectation for each. Record a gate that needs a non-zero expectation in its own artifact file.

A gate whose observed EXIT_CODE equals its declared expectation is normalized to pass. The observed exit code is still displayed, and the rendered row additionally carries - Expected EXIT_CODE: <int> between the EXIT_CODE and Normalized result lines when the expectation is non-zero.

Baseline Evidence Output Summary (Required)

For baseline evidence artifacts stored under evidence/baseline/, include an output summary in addition to the schema fields above:

• Output Summary: <1–20 lines capturing the essential outcome>

The summary should be concise, human-readable, and include the most important result signal (e.g., “All checks passed”, “767 passed”, coverage total, or a brief error description).

If a fail-before run is required but impossible, include a short exception dossier with:

• WhyFailingRunImpossible: <1–3 sentences>
• An alternative proof section (e.g., absence-of-test proof)

Fail-before exception dossiers should be stored under evidence/regression-testing/.

Preferred filename for fail-before exception dossiers:

• fail-before-exception.<timestamp>.md

If an exception dossier is present and schema-valid, it counts as satisfying the "fail-before" requirement for audit/plan reconciliation purposes.

Negative Evidence Claims (Absence Must Be Auditable)

If you claim evidence is missing (e.g., "no exception dossier recorded"), you MUST also record:

• SearchScope: the exact folder(s) searched (include both current-version scope and feature root when applicable)
• SearchPatterns: the filename patterns used (e.g., fail-before-exception.*.md)
• SearchResult: what was found (paths) or none

This prevents false negatives caused by searching the wrong scope or using incomplete patterns.

Evidence-First Audit Writing

When marking FAIL or PARTIAL in audit artifacts, include:

• Concrete file + location (line/hunk/section when possible)
• The violated rule or expected behavior
• The verification command and its output (or why it could not be run)

Issue Update Mirroring (Canonical Location)

When work involves updating a GitHub issue, create a local mirror artifact at:

• <FEATURE>/evidence/issue-updates/issue-<N>.<timestamp>.md

Required contents:

• Timestamp: <ISO-8601>
• The exact text intended/posted
• PostedAs: body or PostedAs: comment (preferred), or PostedAs: unknown
• If posted as a comment: the GitHub URL to the comment
• If posted as an issue body update: the GitHub URL to the issue and IssueUpdatedAt: <ISO-8601>
• If not posted: a POSTING BLOCKED header and the reason

If PostedAs: body, mirror the same update into the local feature issue.md (current version folder if present; otherwise feature root).