Pre-release Review
Run a read-only production release readiness review. The goal is to reduce release time and
coordination failures by finding missing deploy materials, unsafe ordering, configuration gaps,
data migration gaps, and ambiguous production risks before CI/CD or manual release steps begin.
This skill never mutates source code, configs, migrations, secrets, deployment files, or remote
infrastructure. It produces a concise, prioritized report of confirmed problems and plausible
risks that need confirmation.
When to Use
- The user asks for a release audit, pre-release review, go-live review, or deploy readiness check.
- Before publishing a tag, deploying production services, or merging a release branch.
- A PR or git range may include migrations, environment changes, queues, cache behavior, object
storage assets, or service contract changes.
- The user asks whether a change is safe to ship and needs a read-only risk report.
Prerequisites
- Git available on
PATH.
gh CLI (optional) — used only when available and authenticated for PR diffs. If missing or
unauthenticated, fall back to local branch/patch/explicit git range. Do not invent PR contents.
rg (ripgrep) — used for pattern-based evidence collection.
- Windows host (PowerShell) is the primary environment. Commands below use POSIX-style flags
that work in PowerShell when wrapped in double quotes or passed via
git/rg directly. On
Windows, prefer quoting arguments containing <..> ranges with double quotes, e.g.
git diff --stat "v1.2.3..HEAD".
Non-negotiable Rules
- Do not modify source code, configs, migrations, secrets, deployment files, or generated files.
- Do not execute migrations, clear or warm caches, upload assets, trigger CI/CD, deploy
services, publish tags, rotate secrets, or change remote infrastructure.
- Produce a concise report that lists only confirmed problems and plausible risks needing
confirmation. Do not bury the reader in clean checklist items.
- Sort findings from highest to lowest priority (P0 → P1 → P2).
- For each finding include: module, finding, evidence, inferred owner, risk, recommended action.
- Never reveal private keys, account passwords, tokens, certificates, cookies, or full secret
values. Report only: file path, line number, variable name, secret type, and a redacted hint.
- If evidence is incomplete but the risk could block production, list it as a confirmation item.
Required References
Load these reference files at specific points in the workflow:
| Reference file |
When to load |
references/checklist.md |
Before analyzing findings — ensures important release domains (migrations, config, cache, queues, assets, contracts) are not skipped. |
references/report-template.md |
Before writing the final report — keeps priorities, owner inference, secret redaction, and output shape consistent. |
If a reference file is missing, note the limitation in the report's "Unable To Verify" section
only when it affects the release review.
Project Guidance Discovery
Before interpreting the release diff, look for project-local guidance files such as AGENTS.md
and CLAUDE.md in the repository root and relevant service directories.
- Read them when present so the review respects project-specific conventions, service boundaries,
release rules, validation expectations, ownership hints, and known operational constraints.
- Treat project guidance as context for how to interpret risks, not as permission to perform
mutating release actions.
- If project guidance conflicts with this skill's non-negotiable safety rules, the read-only,
no-secret-disclosure rules in this skill win.
- If a relevant guidance file cannot be read, note the limitation in "Unable To Verify" only when
it affects the release review.
Procedure
Step 1 — Determine the Review Scope
State the chosen range in the report.
- PR provided (URL or number): review that PR diff first.
- Explicit
base..head range provided: use it directly.
- Only a head commit provided: compare the previous usable release tag reachable from that
commit to the head commit.
- No scope provided: compare the previous usable release tag to
HEAD.
- Choosing the previous usable release tag:
- Prefer the repository's visible release-tag convention when obvious (semantic versions,
v*, or release-*). If tag naming is mixed, state the assumption.
- If
HEAD is exactly at one or more tags, treat those as the current release point and compare
against the earlier reachable release tag, not HEAD's own tag.
- If no usable previous release tag exists, review the latest 5 commits and explicitly
warn: "No usable previous release tag found; audit only covers the latest 5 commits.
Recommend a PR or tag-based range for future reviews."
Step 2 — Read-Only Evidence Collection
Run only safe inspection commands, adjusted to the repository and current permissions.
git status --short
git rev-parse --show-toplevel
git rev-parse --abbrev-ref HEAD
git rev-parse HEAD
rg --files -g "AGENTS.md" -g "CLAUDE.md"
git tag --merged HEAD --sort=-creatordate
git tag --points-at HEAD
git for-each-ref --sort=-creatordate --format="%(refname:short) %(objectname:short)" refs/tags
git describe --tags --abbrev=0 HEAD
git diff --name-status "<base>..<head>"
git diff --stat "<base>..<head>"
git log --oneline --decorate --no-merges "<base>..<head>"
git diff -U3 "<base>..<head>" -- "<path>"
git blame -L "<start>,<end>" -- "<path>"
git log --format="%h %an %s" -- "<path>"
rg -n "<pattern>" .
For PRs, use gh pr view and gh pr diff only when available and allowed. Do not bypass network,
auth, sandbox, or approval restrictions. If a command cannot run, record the limitation in the
report's "Unable To Verify" section.
Step 3 — Review Workflow
- Confirm the git repository root, current branch, dirty state, and selected comparison range.
- Collect changed file names, file status, diff stats, commit summaries, and touched services.
- Inspect relevant diffs rather than relying on filenames alone.
- Load
references/checklist.md and map changed code to production requirements:
- Schema changes → migrations, indexes, seeds, and backfills.
- Config reads → env examples, deploy secrets, flags, and runtime config.
- Cache key or TTL changes → invalidation, prewarm, and compatibility work.
- Queue producers/consumers → topic setup, DLQ, idempotency, and deploy order.
- Asset references → object storage, CDN, templates, certificates, and permissions.
- Service contract changes → deploy sequence, backward compatibility, and rollback risk.
- Infer owners with
git blame on changed lines when possible; otherwise use recent git log
authors for the file or commit. Label them as inferred owners and do not include email
addresses.
- Classify each finding as P0, P1, or P2 using
references/report-template.md.
- Write the final report in the user's language when practical. Keep conclusion values exactly as
BLOCKED, NEEDS_CONFIRMATION, or NO_BLOCKER_FOUND.
Step 4 — Dirty Worktree Handling
By default, review only the selected committed range. Do not silently mix uncommitted or
untracked changes into the release diff unless the user explicitly asks to include worktree
changes.
- Always report whether the worktree is dirty.
- If dirty or untracked files touch release-relevant areas (migrations, deployment config, env
examples, CI/CD, secrets, cache, queues, assets, or service contracts), add a P2 confirmation
item saying those changes are excluded from the committed-range review and must be committed,
discarded, or reviewed separately before release.
- If the user explicitly asks to include dirty worktree changes, inspect them with read-only
commands such as
git diff and git diff --name-status, and clearly label them as
uncommitted evidence.
Step 5 — Evidence Expectations
Every finding should cite concrete evidence:
- File path and line number when available.
- Commit hash or PR reference when line evidence is not enough.
- Command limitation when evidence could not be collected.
- Diff relationship, e.g. "schema changed but no migration file changed".
Do not state that something is safe just because no file matched a pattern. Use "not verified"
for areas that cannot be confirmed from local repository evidence.
Step 6 — Findings vs. Verification Limits
Separate release confirmation items from neutral tool limits:
- A release confirmation item is a diff-linked production risk — e.g. a new env var whose
production secret cannot be verified, a schema change with unclear migration status, or a new
queue whose infrastructure cannot be confirmed. Classify as P1 or P2 and set conclusion to
NEEDS_CONFIRMATION unless a P0 also exists.
- An "Unable To Verify" entry is a neutral limitation — e.g. missing remote access or
deployment platform credentials when the diff does not introduce a specific release requirement.
Neutral limitations do not change the conclusion by themselves.
- If a limitation blocks confirmation of a release-critical diff change, promote it to a
P1/P2 finding rather than leaving it only in "Unable To Verify".
- Use
NO_BLOCKER_FOUND only when no P0–P2 findings or release confirmation items were found from
available evidence. The report may still include neutral verification limits.
Step 7 — Output Rules
- Show P0 and P1 findings first, then P2 confirmation items.
- Do not list clean checklist categories.
- Include a service deployment order section only when the diff touches multiple services,
asynchronous workers, migrations, queues, cache, or public contracts.
- If no P0 blocker is found but P1/P2 confirmation items remain, use
NEEDS_CONFIRMATION.
- If no P0–P2 findings exist, include the reviewed range and any neutral verification limits.
- Keep the report short enough for a release manager to act on immediately.
Pitfalls
- Mixing uncommitted changes into the release diff — always state whether the worktree is
dirty and exclude uncommitted changes by default.
- Inventing PR contents when
gh is unavailable — never fabricate diff content; ask for a
local branch or explicit range instead.
- Disclosing secrets in the report — report only file path, line number, variable name, secret
type, and a redacted hint. Never include full values.
- Claiming safety from absence of matches — "no file matched a pattern" is not proof of
safety; use "not verified" instead.
- Treating project guidance as permission to mutate —
AGENTS.md/CLAUDE.md are context only;
this skill's read-only rules always win.
- Leaving release-critical limitations as neutral — if a limitation blocks confirmation of a
release-critical diff change, promote it to P1/P2.
- Burying the reader in clean checklist items — the report lists only confirmed problems and
plausible risks, not a full pass/fail checklist.
- Including email addresses in inferred owners — label as inferred owners and omit emails.
Verification
After running the review, verify the skill executed correctly:
Confirm read-only compliance — no files were modified:
git status --short
Expected: same dirty/clean state as before the review (no new changes introduced by the skill).
Confirm scope was stated — the report should explicitly name the reviewed range (e.g.
v1.2.3..HEAD, PR #123, or "latest 5 commits — no usable previous release tag").
Confirm conclusion value — the report's conclusion must be exactly one of:
BLOCKED — a P0 finding exists.
NEEDS_CONFIRMATION — P1/P2 items remain but no P0.
NO_BLOCKER_FOUND — no P0–P2 findings or confirmation items.
Confirm secret redaction — search the report output for any token-like strings:
rg -n "(?i)(api_key|secret|password|token|private_key)\s*[:=]\s*\S+" .
Expected: no full secret values in the report; only redacted hints.
Confirm reference files were loaded — the report should reflect domains from
references/checklist.md and follow the shape from references/report-template.md.
Limitations
- This skill is read-only and does not deploy, tag, publish, run migrations, rotate secrets,
or change infrastructure.
- It can identify release risks from available evidence, but cannot prove production state
without access to the relevant deployment, secrets, database, queue, cache, or observability
systems.
- It should not replace service-owner signoff for high-risk production changes.
Examples
Test Prompts
Use these prompts to validate skill behavior:
- "Run a pre-release review and tell me if this production deploy has risks."
- "Review PR #123 before release. Check migrations, configs, and cache work."
- "This repo has no tags. Use the default strategy and audit release readiness."
- "Check
v1.2.3..HEAD for backend go-live blockers."
Report Conclusion Values
| Conclusion |
Meaning |
BLOCKED |
A P0 finding exists; do not release until resolved. |
NEEDS_CONFIRMATION |
P1/P2 items remain; release may proceed with owner signoff. |
NO_BLOCKER_FOUND |
No P0–P2 findings or confirmation items from available evidence. |
Related Skills
- ci-cd-preflight — validate CI/CD pipeline configuration before triggering a release.
- migration-safety-check — deep-dive into database migration safety and rollback plans.
- secret-scan — scan for leaked secrets in code and configuration files.
1---
2name: pre-release-review
3description: "Runs a read-only go-live review of deploy readiness: migrations, config, secrets redaction, rollout order, rollback risk, and launch blockers, producing a prioritized report. Use when the user asks for a release audit before tagging or deploying. Not for executing migrations, deploying, tagging, or editing release files; never dump live secret values into the report."
4version: 1.0.1
5category: operations
6risk: safe
7source: community
8source_repo: chaunsin/agent-skills
9source_type: community
10date_added: "2026-06-29"
11tags: [release, deploy-readiness, ci-cd, rollback, production]
12tools: [git, gh, rg]
13license: "Apache-2.0"
14license_source: "https://github.com/chaunsin/agent-skills/blob/master/LICENSE"
15metadata:
16 source: https://skillmd.com/skills/kayforkind/pre-release-review
17---
18
19# Pre-release Review
20
21Run a **read-only** production release readiness review. The goal is to reduce release time and
22coordination failures by finding missing deploy materials, unsafe ordering, configuration gaps,
23data migration gaps, and ambiguous production risks **before** CI/CD or manual release steps begin.
24
25This skill never mutates source code, configs, migrations, secrets, deployment files, or remote
26infrastructure. It produces a concise, prioritized report of confirmed problems and plausible
27risks that need confirmation.
28
29## When to Use
30
31- The user asks for a **release audit**, **pre-release review**, **go-live review**, or **deploy readiness check**.
32- Before publishing a tag, deploying production services, or merging a release branch.
33- A PR or git range may include migrations, environment changes, queues, cache behavior, object
34 storage assets, or service contract changes.
35- The user asks whether a change is safe to ship and needs a read-only risk report.
36
37## Prerequisites
38
39- **Git** available on `PATH`.
40- **`gh` CLI** (optional) — used only when available and authenticated for PR diffs. If missing or
41 unauthenticated, fall back to local branch/patch/explicit git range. Do not invent PR contents.
42- **`rg` (ripgrep)** — used for pattern-based evidence collection.
43- **Windows host (PowerShell)** is the primary environment. Commands below use POSIX-style flags
44 that work in PowerShell when wrapped in double quotes or passed via `git`/`rg` directly. On
45 Windows, prefer quoting arguments containing `<..>` ranges with double quotes, e.g.
46 `git diff --stat "v1.2.3..HEAD"`.
47
48## Non-negotiable Rules
49
50- **Do not modify** source code, configs, migrations, secrets, deployment files, or generated files.
51- **Do not execute** migrations, clear or warm caches, upload assets, trigger CI/CD, deploy
52 services, publish tags, rotate secrets, or change remote infrastructure.
53- Produce a **concise report** that lists only confirmed problems and plausible risks needing
54 confirmation. Do not bury the reader in clean checklist items.
55- **Sort findings** from highest to lowest priority (P0 → P1 → P2).
56- For each finding include: **module, finding, evidence, inferred owner, risk, recommended action**.
57- **Never reveal** private keys, account passwords, tokens, certificates, cookies, or full secret
58 values. Report only: file path, line number, variable name, secret type, and a redacted hint.
59- If evidence is incomplete but the risk could block production, list it as a **confirmation item**.
60
61## Required References
62
63Load these reference files at specific points in the workflow:
64
65| Reference file | When to load |
66|---|---|
67| `references/checklist.md` | **Before analyzing findings** — ensures important release domains (migrations, config, cache, queues, assets, contracts) are not skipped. |
68| `references/report-template.md` | **Before writing the final report** — keeps priorities, owner inference, secret redaction, and output shape consistent. |
69
70If a reference file is missing, note the limitation in the report's "Unable To Verify" section
71only when it affects the release review.
72
73## Project Guidance Discovery
74
75Before interpreting the release diff, look for project-local guidance files such as `AGENTS.md`
76and `CLAUDE.md` in the repository root and relevant service directories.
77
78- Read them when present so the review respects project-specific conventions, service boundaries,
79 release rules, validation expectations, ownership hints, and known operational constraints.
80- Treat project guidance as **context** for how to interpret risks, not as permission to perform
81 mutating release actions.
82- If project guidance conflicts with this skill's non-negotiable safety rules, the **read-only,
83 no-secret-disclosure rules in this skill win**.
84- If a relevant guidance file cannot be read, note the limitation in "Unable To Verify" only when
85 it affects the release review.
86
87## Procedure
88
89### Step 1 — Determine the Review Scope
90
91State the chosen range in the report.
92
931. **PR provided** (URL or number): review that PR diff first.
94 - If `gh` is available and authenticated, use read-only commands:
95 ```powershell
96 gh pr view <PR_NUMBER>
97 gh pr diff <PR_NUMBER>
98 ```
99 - If the PR cannot be fetched (missing tooling, auth, or network), say so and ask for a local
100 branch, patch, or explicit git range. **Do not invent the PR contents.**
1012. **Explicit `base..head` range provided**: use it directly.
1023. **Only a head commit provided**: compare the previous usable release tag reachable from that
103 commit to the head commit.
1044. **No scope provided**: compare the previous usable release tag to `HEAD`.
1055. **Choosing the previous usable release tag**:
106 - Prefer the repository's visible release-tag convention when obvious (semantic versions,
107 `v*`, or `release-*`). If tag naming is mixed, state the assumption.
108 - If `HEAD` is exactly at one or more tags, treat those as the current release point and compare
109 against the **earlier** reachable release tag, not `HEAD`'s own tag.
110 - If **no usable previous release tag exists**, review the latest **5 commits** and explicitly
111 warn: *"No usable previous release tag found; audit only covers the latest 5 commits.
112 Recommend a PR or tag-based range for future reviews."*
113
114### Step 2 — Read-Only Evidence Collection
115
116Run only safe inspection commands, adjusted to the repository and current permissions.
117
118```powershell
119git status --short
120git rev-parse --show-toplevel
121git rev-parse --abbrev-ref HEAD
122git rev-parse HEAD
123rg --files -g "AGENTS.md" -g "CLAUDE.md"
124git tag --merged HEAD --sort=-creatordate
125git tag --points-at HEAD
126git for-each-ref --sort=-creatordate --format="%(refname:short) %(objectname:short)" refs/tags
127git describe --tags --abbrev=0 HEAD
128git diff --name-status "<base>..<head>"
129git diff --stat "<base>..<head>"
130git log --oneline --decorate --no-merges "<base>..<head>"
131git diff -U3 "<base>..<head>" -- "<path>"
132git blame -L "<start>,<end>" -- "<path>"
133git log --format="%h %an %s" -- "<path>"
134rg -n "<pattern>" .
135```
136
137For PRs, use `gh pr view` and `gh pr diff` only when available and allowed. Do not bypass network,
138auth, sandbox, or approval restrictions. If a command cannot run, record the limitation in the
139report's "Unable To Verify" section.
140
141### Step 3 — Review Workflow
142
1431. Confirm the git repository root, current branch, dirty state, and selected comparison range.
1442. Collect changed file names, file status, diff stats, commit summaries, and touched services.
1453. Inspect relevant diffs rather than relying on filenames alone.
1464. Load `references/checklist.md` and map changed code to production requirements:
147 - **Schema changes** → migrations, indexes, seeds, and backfills.
148 - **Config reads** → env examples, deploy secrets, flags, and runtime config.
149 - **Cache key or TTL changes** → invalidation, prewarm, and compatibility work.
150 - **Queue producers/consumers** → topic setup, DLQ, idempotency, and deploy order.
151 - **Asset references** → object storage, CDN, templates, certificates, and permissions.
152 - **Service contract changes** → deploy sequence, backward compatibility, and rollback risk.
1535. Infer owners with `git blame` on changed lines when possible; otherwise use recent `git log`
154 authors for the file or commit. Label them as **inferred owners** and **do not include email
155 addresses**.
1566. Classify each finding as **P0**, **P1**, or **P2** using `references/report-template.md`.
1577. Write the final report in the user's language when practical. Keep conclusion values exactly as
158 `BLOCKED`, `NEEDS_CONFIRMATION`, or `NO_BLOCKER_FOUND`.
159
160### Step 4 — Dirty Worktree Handling
161
162By default, review only the selected committed range. Do not silently mix uncommitted or
163untracked changes into the release diff unless the user explicitly asks to include worktree
164changes.
165
166- **Always report** whether the worktree is dirty.
167- If dirty or untracked files touch release-relevant areas (migrations, deployment config, env
168 examples, CI/CD, secrets, cache, queues, assets, or service contracts), add a **P2 confirmation
169 item** saying those changes are excluded from the committed-range review and must be committed,
170 discarded, or reviewed separately before release.
171- If the user explicitly asks to include dirty worktree changes, inspect them with read-only
172 commands such as `git diff` and `git diff --name-status`, and clearly label them as
173 **uncommitted evidence**.
174
175### Step 5 — Evidence Expectations
176
177Every finding should cite concrete evidence:
178
179- File path and line number when available.
180- Commit hash or PR reference when line evidence is not enough.
181- Command limitation when evidence could not be collected.
182- Diff relationship, e.g. *"schema changed but no migration file changed"*.
183
184Do not state that something is safe just because no file matched a pattern. Use **"not verified"**
185for areas that cannot be confirmed from local repository evidence.
186
187### Step 6 — Findings vs. Verification Limits
188
189Separate release confirmation items from neutral tool limits:
190
191- A **release confirmation item** is a diff-linked production risk — e.g. a new env var whose
192 production secret cannot be verified, a schema change with unclear migration status, or a new
193 queue whose infrastructure cannot be confirmed. Classify as P1 or P2 and set conclusion to
194 `NEEDS_CONFIRMATION` unless a P0 also exists.
195- An **"Unable To Verify"** entry is a neutral limitation — e.g. missing remote access or
196 deployment platform credentials when the diff does not introduce a specific release requirement.
197 Neutral limitations do not change the conclusion by themselves.
198- If a limitation blocks confirmation of a release-critical diff change, **promote it** to a
199 P1/P2 finding rather than leaving it only in "Unable To Verify".
200- Use `NO_BLOCKER_FOUND` only when no P0–P2 findings or release confirmation items were found from
201 available evidence. The report may still include neutral verification limits.
202
203### Step 7 — Output Rules
204
205- Show **P0 and P1** findings first, then **P2** confirmation items.
206- **Do not list** clean checklist categories.
207- Include a **service deployment order** section only when the diff touches multiple services,
208 asynchronous workers, migrations, queues, cache, or public contracts.
209- If no P0 blocker is found but P1/P2 confirmation items remain, use `NEEDS_CONFIRMATION`.
210- If no P0–P2 findings exist, include the reviewed range and any neutral verification limits.
211- Keep the report short enough for a release manager to act on immediately.
212
213## Pitfalls
214
215- **Mixing uncommitted changes into the release diff** — always state whether the worktree is
216 dirty and exclude uncommitted changes by default.
217- **Inventing PR contents** when `gh` is unavailable — never fabricate diff content; ask for a
218 local branch or explicit range instead.
219- **Disclosing secrets in the report** — report only file path, line number, variable name, secret
220 type, and a redacted hint. Never include full values.
221- **Claiming safety from absence of matches** — "no file matched a pattern" is not proof of
222 safety; use "not verified" instead.
223- **Treating project guidance as permission to mutate** — `AGENTS.md`/`CLAUDE.md` are context only;
224 this skill's read-only rules always win.
225- **Leaving release-critical limitations as neutral** — if a limitation blocks confirmation of a
226 release-critical diff change, promote it to P1/P2.
227- **Burying the reader in clean checklist items** — the report lists only confirmed problems and
228 plausible risks, not a full pass/fail checklist.
229- **Including email addresses in inferred owners** — label as inferred owners and omit emails.
230
231## Verification
232
233After running the review, verify the skill executed correctly:
234
2351. **Confirm read-only compliance** — no files were modified:
236 ```powershell
237 git status --short
238 ```
239 Expected: same dirty/clean state as before the review (no new changes introduced by the skill).
240
2412. **Confirm scope was stated** — the report should explicitly name the reviewed range (e.g.
242 `v1.2.3..HEAD`, PR #123, or "latest 5 commits — no usable previous release tag").
243
2443. **Confirm conclusion value** — the report's conclusion must be exactly one of:
245 - `BLOCKED` — a P0 finding exists.
246 - `NEEDS_CONFIRMATION` — P1/P2 items remain but no P0.
247 - `NO_BLOCKER_FOUND` — no P0–P2 findings or confirmation items.
248
2494. **Confirm secret redaction** — search the report output for any token-like strings:
250 ```powershell
251 rg -n "(?i)(api_key|secret|password|token|private_key)\s*[:=]\s*\S+" .
252 ```
253 Expected: no full secret values in the report; only redacted hints.
254
2555. **Confirm reference files were loaded** — the report should reflect domains from
256 `references/checklist.md` and follow the shape from `references/report-template.md`.
257
258## Limitations
259
260- This skill is **read-only** and does not deploy, tag, publish, run migrations, rotate secrets,
261 or change infrastructure.
262- It can identify release risks from available evidence, but **cannot prove production state**
263 without access to the relevant deployment, secrets, database, queue, cache, or observability
264 systems.
265- It should **not replace service-owner signoff** for high-risk production changes.
266
267## Examples
268
269### Test Prompts
270
271Use these prompts to validate skill behavior:
272
273- *"Run a pre-release review and tell me if this production deploy has risks."*
274- *"Review PR #123 before release. Check migrations, configs, and cache work."*
275- *"This repo has no tags. Use the default strategy and audit release readiness."*
276- *"Check `v1.2.3..HEAD` for backend go-live blockers."*
277
278### Report Conclusion Values
279
280| Conclusion | Meaning |
281|---|---|
282| `BLOCKED` | A P0 finding exists; do not release until resolved. |
283| `NEEDS_CONFIRMATION` | P1/P2 items remain; release may proceed with owner signoff. |
284| `NO_BLOCKER_FOUND` | No P0–P2 findings or confirmation items from available evidence. |
285
286## Related Skills
287
288- **ci-cd-preflight** — validate CI/CD pipeline configuration before triggering a release.
289- **migration-safety-check** — deep-dive into database migration safety and rollback plans.
290- **secret-scan** — scan for leaked secrets in code and configuration files.