--- title: "Trail of Bits Differential Security Review: review, setup & limitations · Undominated.ai" canonical: https://undominated.ai/skills/trailofbits-differential-review/ description: "Reviews security-relevant changes against their baseline, removed protections, callers and concrete attacker scenarios." --- # Trail of Bits Differential Security Review: review, setup & limitations · Undominated.ai > Reviews security-relevant changes against their baseline, removed protections, callers and concrete attacker scenarios. [← Explore all skills](/skills/) SECURITY / Trail of Bits # Trail of Bits Differential Security Review Reviews security-relevant changes against their baseline, removed protections, callers and concrete attacker scenarios. See setup guidance ↓Original source ↗ SOURCE REVIEW Reviewed 2026-09-21 Evidence 7 linked sources Publisher Trail of Bits Licence CC-BY-SA-4.0 ↗ Revision 123037ec8aed Read what was—and wasn’t—checked ↓ “Performs security-focused differential review of code changes. Adapts analysis depth to codebase size, uses git blame for context, calculates blast radius by counting callers, checks” Trail of Bits · upstream description ↗ Our analysis follows below. 01 / THE REASONING ## Why this made the selection. - Uses git history to ask why a removed check existed and whether a previous fix is being undone. - Requires findings tied to lines and commits and a saved report with stated coverage limits. ### A good fit for - Reviewing a pull request that changes authorization or data validation - Investigating security regressions in a patch ### Weigh up before choosing - The methodology includes checking out the baseline and head; use a suitable worktree so the review does not disrupt uncommitted work. - Its grep-based caller counts and test searches are heuristics, not a resolved call graph or measured test coverage. - Risk-based triage can leave lower-priority paths unexamined; the final coverage statement matters. 02 / THE REVIEW RECORD ## What we actually inspected. Source review has boundaries. A clear record is more useful than a “safe” badge. ### Material inspected - plugins/differential-review/skills/differential-review/SKILL.md - plugins/differential-review/skills/differential-review/methodology.md - plugins/differential-review/skills/differential-review/reporting.md - plugins/differential-review/agents/adversarial-modeler.md - LICENSE ### Our findings - The inspected methodology connects baseline inspection, changed-code review, test evidence and transitive effects. - High-risk work can use the plugin’s namespaced adversarial-modeler agent, which needs the full plugin. - The reporting guide asks for exploit prerequisites, impact and evidence rather than a bare severity label. ### Not established by this review - Source inspection only; no skill installation or workflow execution was performed. The review applies to the material and revision named here. A newer upstream release can change its behavior. 03 / PUT IT TO WORK ## Add a skill to your workflow. Upstream setup instructions ↗ - Install the complete differential-review plugin, including its agents, workflows and references. - Claude Code: add the trailofbits/skills marketplace with /plugin marketplace add trailofbits/skills, then select the plugin in /plugin menu. - The publisher also documents Codex marketplace compatibility; confirm the required workflow and delegation capabilities in the chosen host. ### Before you start - A supported Claude Code or Codex plugin host - Permission for the plugin’s repository reads, outputs and any requested test commands - A baseline revision and an appropriate clean or isolated working tree ### Compatibility Claude Code and documented Codex plugin compatibility · Git repositories with a meaningful baseline ### Read diffs, history, callers and tests - Read diffs, history, callers and tests - Run git commands that can change the checked-out revision - Write review reports and optionally run authorized validation ### Cost model The package uses CC-BY-SA terms; agent review and any validation infrastructure are separate costs. 04 / FOLLOW THE EVIDENCE ## The source trail. Our notes are separate from the original resource. Check upstream before adopting a new version. - Skill definition at reviewed revision ↗ Checked 2026-09-21 https://github.com/trailofbits/skills/blob/123037ec8aed26f0d86327cc39137ee5043e5deb/plugins/differential-review/skills/differential-review/SKILL.md Supports: summary, upstreamDescription, whySelected, bestFor, limitations, review, compatibility, access - Reviewed reference or helper: methodology.md ↗ Checked 2026-09-21 https://github.com/trailofbits/skills/blob/123037ec8aed26f0d86327cc39137ee5043e5deb/plugins/differential-review/skills/differential-review/methodology.md Supports: review, limitations, access - Reviewed reference or helper: reporting.md ↗ Checked 2026-09-21 https://github.com/trailofbits/skills/blob/123037ec8aed26f0d86327cc39137ee5043e5deb/plugins/differential-review/skills/differential-review/reporting.md Supports: review, limitations, access - Reviewed reference or helper: ../../agents/adversarial-modeler.md ↗ Checked 2026-09-21 https://github.com/trailofbits/skills/blob/123037ec8aed26f0d86327cc39137ee5043e5deb/plugins/differential-review/agents/adversarial-modeler.md Supports: review, limitations, access - Upstream license or licensing declaration ↗ Checked 2026-09-21 https://github.com/trailofbits/skills/blob/123037ec8aed26f0d86327cc39137ee5043e5deb/LICENSE Supports: license - Publisher installation and project documentation ↗ Checked 2026-09-21 https://github.com/trailofbits/skills/blob/123037ec8aed26f0d86327cc39137ee5043e5deb/README.md Supports: install, compatibility, access - Skills installer source formats and selective installation ↗ Checked 2026-09-21 https://github.com/vercel-labs/skills/blob/7407f3893ad4dceab546ac002c3ef806e4000c73/README.md Supports: install KEEP COMPARING ## Other approaches to consider. Related by category or shared topics. These are alternatives to inspect, not a measured quality order. [### OpenAI Codex Security Threat Model ↗ Creates or reuses a repository threat model bound to the actual target and revision, preserving supplied context and explicit review gaps.](/skills/openai-threat-model/)[### Trail of Bits Audit Context Building ↗ Builds a cited dossier of function invariants, assumptions and dependencies before a separate vulnerability-hunting pass.](/skills/trailofbits-audit-context-building/)[### Trail of Bits Sharp Edges ↗ Examines APIs and configuration for cases where ordinary developer mistakes can disable a security boundary.](/skills/trailofbits-sharp-edges/) [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)