Install
$ agentstack add skill-lakesoftai-deepdone-agent-skills-deepdone-review ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
DeepDone Review
Purpose
Review local diff like hostile teammate who wants truth.
Primary Review Axes
- correctness
- regressions
- security and permissions
- overscoped changes
- missing or weak tests
- rollback pain
Severity Rubric
Use these labels:
block: must fix before integration. Includes correctness break, security exposure, data loss risk, broken migration, failing required check, or behavior that contradicts the milestone.strong-concern: likely should fix before commit, but may need user prioritization. Includes brittle design, weak verification on risky paths, broad coupling, or hidden rollback cost.nit: small local cleanup that does not change safety or correctness.
Do not inflate style issues into blockers. Do not bury real blockers as nits.
Review Behavior
- findings first
- severity order
- cite exact files and lines when possible
- no praise padding
- if no findings, say so plainly and call out remaining test gaps or uncertainty
- include clear fix direction for each actionable finding
- separate product or architecture questions from code findings
Ledger Writeback
When epic ledger exists:
- summarize accepted findings or clean review result under
## Decisionsor## Open Loops - include one structured line:
review-result: pass|fail|blocked - note residual risks plainly
- update
Statusif review blocks integration
Set review-result as:
pass: no blocking or strong unresolved findingsfail: review found local defects that should be fixed before commitblocked: review found product, architecture, security, migration, or prioritization judgment that needs user input
Surface Checklists
Auth And Permissions
- caller identity is verified at the correct boundary
- authorization checks happen before data access or mutation
- failures do not leak sensitive internal state
- tests include allowed and denied paths
Schema And Migrations
- migration is reversible or rollback pain is explicit
- reads and writes work across expected old and new shapes
- nullable, default, and backfill behavior is deliberate
- migration ordering is safe for deploy reality
Config And Deployment
- defaults are safe
- secrets are not hardcoded or logged
- environment-specific behavior is explicit
- generated or local-only files are not included
Public API Or User-Facing Behavior
- response shape, status code, event name, or UI state matches contract
- error cases are covered, not only happy path
- compatibility impact is stated
- tests prove the user-visible acceptance condition
Shared Libraries
- callers outside the milestone still satisfy the new contract
- behavior changes are named in tests or docs
- narrow fix did not become broad refactor
Workflow
- Read diff and recent verification results.
- Check whether implementation matches stated milestone.
- Look for behavior regressions and hidden coupling.
- Inspect auth, data, config, and migration surfaces with extra skepticism.
- Check whether tests prove intended behavior.
- Classify each finding by severity.
- Write
review-resultinto the active ledger when one exists. - Report findings first.
Output
Return:
- findings ordered by severity
- open questions or assumptions
- brief integration readiness statement
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: lakesoftai
- Source: lakesoftai/deepdone-agent-skills
- License: Apache-2.0
- Homepage: https://deepdone.ai
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.