Change Verdict
Use the smallest mode or action and scope that answer the request. The six
modes, pending, and default deps read evidence and return results without
changing workspace or review state. Explicit dependency-remediation authority
permits only a bounded deps fix delegation to an available authorized writer.
The parent retains its read-only command grants; neither action metadata nor
the user's fix request adds direct shell, Write, or Edit capability.
Modes
| Mode |
Question answered |
Read when |
code |
Is the selected code change sound against applicable standards and specifications? |
Reviewing a code diff or branch. |
pr |
Is a proposed PR complete and hygienic for its declared merge method? |
Reviewing branch changes or PR metadata. |
security |
Are security-sensitive paths exposed to OWASP risks or unsafe dependencies? |
Auditing auth, input, secrets, dependencies, or security-sensitive changes. |
tests |
Do existing tests and acceptance evidence cover the behavior? |
Reviewing test adequacy or tracing acceptance criteria to evidence. |
docs |
Is a document accurate, complete, and consistent with the code? |
Reviewing Markdown, specs, READMEs, or design documents. |
risk |
What breaking surface, blast radius, and change-scope signals are present? |
Assessing an uncommitted diff or large refactor. |
Choose one mode. If the mode or requested scope is ambiguous or cannot be
resolved, return INCONCLUSIVE and state the minimum evidence or clarification
needed.
Actions
Natural-language pending-review and dependency requests select these actions;
selectors are optional when the task is clear. Select one action or mode per
observation and load only its local procedure.
| Action |
Procedure and authority |
pending [--files "<rel-path,...>"] |
Read references/actions/pending.md. Delegate read-only to the available code-reviewer and relay its report. Missing reviewer returns UNAVAILABLE, with no substitute. |
deps [--level <severity>] |
Read references/actions/deps.md. Use an available read-only audit capability and obtain a separate independent security verdict. |
deps fix [--level <severity>] or deps --fix |
Read the same dependency procedure. Delegate only the explicitly authorized dependency scope to an available writer, followed by an independent read-only security verdict. Missing bounded writer capability returns UNAVAILABLE; reuse existing authority for that scope. |
When NOT to Use
- Implement or fix a source finding: use
flow-drive. Explicit dependency
remediation belongs to deps fix within its declared bound.
- Trace an unfamiliar code path: use
code-trace.
- Specialized skill governance scoring or stocktake: the former audit capability is retired. For an existing skill document or change, select this skill's
docs or risk mode within its declared scope.
Procedure
Follow references/shared/review-workflow.md
for scope resolution, fixed-point handling, evidence collection, and aggregation.
It owns the shared read-only sequence.
Load only the selected mode's references:
code: use the branch prompt for a branch scope; otherwise use the fast or
full prompt matching the selected depth. The shared code research instructions
own Git reads; apply the code-only Standards and Spec status rules in
references/shared/review-common.md.
pr: read PR metadata and use bash ${CLAUDE_SKILL_DIR}/scripts/check-unrelated-changes.sh only as a
read-only advisory scan when PR metadata is available.
security: use references/security/codex-prompt-security.md and label
findings with their OWASP category.
tests: read source, tests, acceptance criteria, and available runtime
evidence. --ac-trace maps criteria to evidence; --coverage loads
references/tests/codex-prompt-test-review.md. Legacy --scope tests means
--mode tests.
docs: read the full target and enough source or configuration to check it.
Load references/docs/review-loop-doc.md only when a prior snapshot is
supplied.
risk: run node ${CLAUDE_SKILL_DIR}/scripts/risk-analyze.js --json on the current read-only tree and
use references/risk/ to interpret its score.
${CLAUDE_SKILL_DIR} is the directory containing this SKILL.md. Claude substitutes it in these helper commands and in allowed-tools. Hosts that do not substitute it, and references that write $SKILL_DIR, resolve that same directory. The working
directory stays the consumer repository. risk-analyze.js and check-unrelated-changes.sh are the helper paths granted by allowed-tools; the grant is not unrestricted Bash. codex review and review-cli.sh stay explicit opt-in and do not start by themselves.
For every mode, report evidence gaps and use the response shape in
templates/review_output.md. The code-only
Standards and Spec section is omitted from all other modes. Severity and final
verdict meanings are owned by
references/shared/review-rubric.md.
Optional CLI second opinion
Only explicit --second-opinion=codex-exec may invoke
bash ${CLAUDE_SKILL_DIR}/scripts/review-cli.sh --backend cli. Pass the selected scope and fixed point,
not the primary conclusion. Keep the CLI result separate, redact it, and
record its status. The primary review is complete without this optional check;
an unavailable CLI is reported as degraded evidence.
Authority boundary
Return review results in the response. The six modes, pending, and default
deps keep tracked/untracked files, index, and lockfiles unchanged. Their
read-only scope excludes fixers, generators, staging, commits, artifact
writers, gate/sentinel emitters, and writer dispatch. deps fix is the sole
bounded writer-delegation exception described above; fix success cannot clear a gate or
replace the independent security verdict. A re-review is a new observation
of a caller-supplied snapshot. Missing required local resources fail closed
before dependent action; obtain no replacement from sibling skills or an
ambient source checkout.
References
references/shared/review-workflow.md — shared sequence and fixed-point handling.
references/shared/review-common.md — evidence, source labels, and finding normalization.
references/shared/review-rubric.md — severity and final verdict definitions.
references/shared/codex-research-instructions.md — research guidance used by code prompts.
references/shared/cli-backend.md — explicit CLI transport contract.
references/actions/pending.md, references/actions/deps.md — delegated
pending review and dependency audit/remediation procedures.
references/code/, references/security/, references/tests/, references/docs/, references/risk/ — mode-specific instructions.
scripts/review-cli.sh, scripts/check-unrelated-changes.sh, scripts/risk-analyze.js — read-only helpers.
Completion check
- One mode or action and a readable, resolved scope are recorded.
- The reported fixed point and evidence anchors match the reviewed snapshot.
- Findings are normalized and ranked; missing evidence remains visible.
- Only an explicitly requested CLI opinion is reported, with its actual status.
- Read-only observations leave repository state unchanged; an explicit
dependency fix records the authorized writer, bounded changes, and a
separate independent security result. No review artifact, gate, or
sentinel was written.
1---2name: change-verdict3description: 'Review code, PRs, security, tests, documents, change risk, pending changes, or dependency vulnerabilities. Verdicts are read-only; explicit dependency fixes go to an authorized writer. Not for source repair or architecture-only design. Output: evidence-backed findings and review status.'4allowed-tools: 'Bash(git diff:*), Bash(git status:*), Bash(git merge-base:*), Bash(git rev-parse:*), Bash(git show:*), Bash(git log:*), Bash(git ls-files:*), Bash(git hash-object:*), Bash(gh pr view:*), Bash(node ${CLAUDE_SKILL_DIR}/scripts/risk-analyze.js:*), Bash(bash ${CLAUDE_SKILL_DIR}/scripts/check-unrelated-changes.sh:*), Read, Grep, Glob, Agent'5metadata:6 source: https://skillmd.com/skills/hmj1026/change-verdict7 dhpk-capability-id: dhpk.skill.change-verdict8 dhpk-name-style: portable-family9 dhpk-invocation-class: implicit-eligible10 dhpk-modes: [code, pr, security, tests, docs, risk]11 dhpk-next:12 PASS: none13 FAIL: none14 BLOCKED: none15 NOT_RUN: none16 UNAVAILABLE: none17 code:PASS: precommit18 docs:PASS: git-smart-commit19---2021# Change Verdict2223Use the smallest mode or action and scope that answer the request. The six24modes, `pending`, and default `deps` read evidence and return results without25changing workspace or review state. Explicit dependency-remediation authority26permits only a bounded `deps fix` delegation to an available authorized writer.27The parent retains its read-only command grants; neither action metadata nor28the user's fix request adds direct shell, `Write`, or `Edit` capability.2930## Modes3132| Mode | Question answered | Read when |33|---|---|---|34| `code` | Is the selected code change sound against applicable standards and specifications? | Reviewing a code diff or branch. |35| `pr` | Is a proposed PR complete and hygienic for its declared merge method? | Reviewing branch changes or PR metadata. |36| `security` | Are security-sensitive paths exposed to OWASP risks or unsafe dependencies? | Auditing auth, input, secrets, dependencies, or security-sensitive changes. |37| `tests` | Do existing tests and acceptance evidence cover the behavior? | Reviewing test adequacy or tracing acceptance criteria to evidence. |38| `docs` | Is a document accurate, complete, and consistent with the code? | Reviewing Markdown, specs, READMEs, or design documents. |39| `risk` | What breaking surface, blast radius, and change-scope signals are present? | Assessing an uncommitted diff or large refactor. |4041Choose one mode. If the mode or requested scope is ambiguous or cannot be42resolved, return `INCONCLUSIVE` and state the minimum evidence or clarification43needed.4445## Actions4647Natural-language pending-review and dependency requests select these actions;48selectors are optional when the task is clear. Select one action or mode per49observation and load only its local procedure.5051| Action | Procedure and authority |52|---|---|53| `pending [--files "<rel-path,...>"]` | Read [`references/actions/pending.md`](references/actions/pending.md). Delegate read-only to the available `code-reviewer` and relay its report. Missing reviewer returns `UNAVAILABLE`, with no substitute. |54| `deps [--level <severity>]` | Read [`references/actions/deps.md`](references/actions/deps.md). Use an available read-only audit capability and obtain a separate independent security verdict. |55| `deps fix [--level <severity>]` or `deps --fix` | Read the same dependency procedure. Delegate only the explicitly authorized dependency scope to an available writer, followed by an independent read-only security verdict. Missing bounded writer capability returns `UNAVAILABLE`; reuse existing authority for that scope. |5657## When NOT to Use5859- Implement or fix a source finding: use `flow-drive`. Explicit dependency60 remediation belongs to `deps fix` within its declared bound.61- Trace an unfamiliar code path: use `code-trace`.62- Specialized skill governance scoring or stocktake: the former audit capability is retired. For an existing skill document or change, select this skill's `docs` or `risk` mode within its declared scope.6364## Procedure6566Follow [`references/shared/review-workflow.md`](references/shared/review-workflow.md)67for scope resolution, fixed-point handling, evidence collection, and aggregation.68It owns the shared read-only sequence.6970Load only the selected mode's references:7172- `code`: use the branch prompt for a branch scope; otherwise use the fast or73 full prompt matching the selected depth. The shared code research instructions74 own Git reads; apply the code-only Standards and Spec status rules in75 [`references/shared/review-common.md`](references/shared/review-common.md).76- `pr`: read PR metadata and use `bash ${CLAUDE_SKILL_DIR}/scripts/check-unrelated-changes.sh` only as a77 read-only advisory scan when PR metadata is available.78- `security`: use `references/security/codex-prompt-security.md` and label79 findings with their OWASP category.80- `tests`: read source, tests, acceptance criteria, and available runtime81 evidence. `--ac-trace` maps criteria to evidence; `--coverage` loads82 `references/tests/codex-prompt-test-review.md`. Legacy `--scope tests` means83 `--mode tests`.84- `docs`: read the full target and enough source or configuration to check it.85 Load `references/docs/review-loop-doc.md` only when a prior snapshot is86 supplied.87- `risk`: run `node ${CLAUDE_SKILL_DIR}/scripts/risk-analyze.js --json` on the current read-only tree and88 use `references/risk/` to interpret its score.8990`${CLAUDE_SKILL_DIR}` is the directory containing this `SKILL.md`. Claude substitutes it in these helper commands and in `allowed-tools`. Hosts that do not substitute it, and references that write `$SKILL_DIR`, resolve that same directory. The working91directory stays the consumer repository. `risk-analyze.js` and `check-unrelated-changes.sh` are the helper paths granted by `allowed-tools`; the grant is not unrestricted Bash. `codex review` and `review-cli.sh` stay explicit opt-in and do not start by themselves.9293For every mode, report evidence gaps and use the response shape in94[`templates/review_output.md`](templates/review_output.md). The code-only95Standards and Spec section is omitted from all other modes. Severity and final96verdict meanings are owned by97[`references/shared/review-rubric.md`](references/shared/review-rubric.md).9899## Optional CLI second opinion100101Only explicit `--second-opinion=codex-exec` may invoke102`bash ${CLAUDE_SKILL_DIR}/scripts/review-cli.sh --backend cli`. Pass the selected scope and fixed point,103not the primary conclusion. Keep the CLI result separate, redact it, and104record its status. The primary review is complete without this optional check;105an unavailable CLI is reported as degraded evidence.106107## Authority boundary108109Return review results in the response. The six modes, `pending`, and default110`deps` keep tracked/untracked files, index, and lockfiles unchanged. Their111read-only scope excludes fixers, generators, staging, commits, artifact112writers, gate/sentinel emitters, and writer dispatch. `deps fix` is the sole113bounded writer-delegation exception described above; fix success cannot clear a gate or114replace the independent security verdict. A re-review is a new observation115of a caller-supplied snapshot. Missing required local resources fail closed116before dependent action; obtain no replacement from sibling skills or an117ambient source checkout.118119## References120121- `references/shared/review-workflow.md` — shared sequence and fixed-point handling.122- `references/shared/review-common.md` — evidence, source labels, and finding normalization.123- `references/shared/review-rubric.md` — severity and final verdict definitions.124- `references/shared/codex-research-instructions.md` — research guidance used by code prompts.125- `references/shared/cli-backend.md` — explicit CLI transport contract.126- `references/actions/pending.md`, `references/actions/deps.md` — delegated127 pending review and dependency audit/remediation procedures.128- `references/code/`, `references/security/`, `references/tests/`, `references/docs/`, `references/risk/` — mode-specific instructions.129- `scripts/review-cli.sh`, `scripts/check-unrelated-changes.sh`, `scripts/risk-analyze.js` — read-only helpers.130131## Completion check132133- One mode or action and a readable, resolved scope are recorded.134- The reported fixed point and evidence anchors match the reviewed snapshot.135- Findings are normalized and ranked; missing evidence remains visible.136- Only an explicitly requested CLI opinion is reported, with its actual status.137- Read-only observations leave repository state unchanged; an explicit138 dependency fix records the authorized writer, bounded changes, and a139 separate independent security result. No review artifact, gate, or140 sentinel was written.