Terminal
Terminal/Skill
Pre release readiness review of a Forge app covering manifest wiring, dependencies, tests and obvious security or cost smells.
By Atlassian
Works with
npx skills add atlassian/forge-skills --skill forge-app-reviewRun a general Forge release-readiness review. This skill is the front door for broad app review, not a replacement for specialist security, cost, or debugging skills.
Use this skill for:
Use another skill instead when the user's primary intent is:
forge-security-review.forge-cost-optimizer.forge-debugger.If a broad review finds a deep security/cost/debug concern, include it as a handoff recommendation rather than duplicating the specialist workflow.
Detect modules declared in manifest.yml or package dependencies, and load specific review guides:
graph:connector, teamwork-graph-connector, or imports @forge/teamwork-graph:./modules/connector-review.md.manifest.yml or manifest.yaml.
package.json.
resolver.define, handler exports, product API calls, storage usage, external fetches, logging, error handling.invoke() patterns, bridge usage, loading/error states.resolver.define() names.forge lint, build, deploy, or install.Only flag obvious signals and recommend forge-security-review for deep validation:
api.asApp() in user-triggered resolvers without obvious authorization checks.Only flag obvious signals and recommend forge-cost-optimizer for deep analysis:
invoke() calls on page load.ignoreSelf where applicable.Only flag readiness gaps; use forge-debugger when there is an observed failure:
forge lint.Return a concise Markdown report. Findings must be a table (not a numbered list). Include a Source column for every finding so the reader knows which review guide or area produced it.
Source values (use the most specific that applies — Source is the checklist/module or review area, not merely the file type):
manifest — general Forge manifest wiring, scopes, egress, or module keys not covered by a module-specific guideresolver — backend/resolver wiring and runtime behaviorfrontend — UI Kit / Custom UI invoke and bridge patternsdependencies — package.json / runtime package fittests — missing or inadequate verificationgeneral — cross-cutting readiness hygiene not covered above./modules/), use the Source label that guide defines (for example connector). Do not re-label those findings as manifest just because evidence lives in manifest.yml.Location rules (do not path-only):
path:line or path:start-end (multiple citations OK).``` fences reliably, wrap the excerpt in HTML: <pre><code>...</code></pre>.# Forge App Review Results
## Summary
- Readiness: Ready | Needs changes | Blocked
- Highest-risk area: <manifest | resolver wiring | permissions | dependencies | tests | operational hygiene | module-specific>
- Files inspected: <short list>
- Specialist handoffs: <none | security | cost | debugger>
## Findings
| Severity | Source | Finding | Location | Description | Doc | Fix |
|----------|--------|---------|----------|-------------|---------|-----|
| Critical \| Warning \| Info | <source> | <short title> | `path:start-end`<br><pre><code>…excerpt…</code></pre> | <why this matters / observed pattern> | <DAC or checklist anchor title + URL> | <specific remediation> |
Sort rows Critical → Warning → Info. Omit the Doc column cell only when no public/doc anchor applies; keep the column.
## Clean Areas
- <important categories checked with no issues>
## Suggested Next Step
- <apply fixes | run specialist review | deploy/lint/test command>
If there are no findings, say the app looks ready from this general review and list any residual specialist reviews that were intentionally out of scope.