review-guidelines

Repository-specific Cerebro review invariants and triage guidance for Droid code and security reviews.

Quality

51%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./.factory/skills/review-guidelines/SKILL.md

Evaluation results

85%

Greenhouse Connector Code Review

Source connector HTTP safety review

Criteria

Without context

With context

Flags connector-local http.Client

80%

Cites sourcehttp invariant

20%

70%

Flags bare io.ReadAll on success body

100%

Cites LimitReader invariant

100%

Mentions preflight command

Structured by pass

100%

Each finding labels its pass

100%

Review memory referenced

100%

Prioritizes security bugs over style

100%

Scoped to changed code

100%

87%

88%

-4%

Graph Ask Query Review

Graph Ask Cypher query safety review

Criteria

Without context

With context

Flags string interpolation for tenantID

100%

Cites tenant-scoping invariant

100%

Flags post-processing on LLM rows

100%

Cites LLM post-processing invariant

100%

Flags no row limit on freeform path

20%

References review memory

100%

Structured by named passes

100%

Each finding labels its pass

100%

Does not request unnecessary tests

100%

75%

Scoped to changed code

100%

99%

-1%

Auth and ATS State Machine Review

Device auth and ATS state transition review

Criteria

Without context

With context

Flags inline client IP derivation

100%

Cites request-origin helper invariant

100%

Flags inline DPoP htu construction

100%

Cites DPoP invariant

100%

Flags split read-then-write in AdvanceToInterview

100%

Flags split read-then-write in RejectCandidate

100%

90%

Cites atomicity invariant

100%

References review memory

100%

Structured by named passes

100%

Findings label their pass

100%

Repository: writer/cerebro
Commit: 3aeaf20

Evaluated: 3 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Greenhouse Connector Code Review Graph Ask Query Review Auth and ATS State Machine Review

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.