Thermos
- 602 installs
- 2.5k repo stars
- Updated August 5, 2026
- cursor/plugins
thermos is a Cursor plugin skill that launches two thermo-nuclear review subagents in parallel and synthesizes bug, security, and code-quality findings for developers auditing medium-risk pull requests without the full e
About
thermos is a Cursor plugins skill that runs condensed thermo-nuclear branch review for medium-risk PRs. The workflow scopes changed files, gathers `git diff` context, and launches two background subagents in one message: `thermo-nuclear-review-subagent` for bugs, breakages, security, devex regressions, and feature-flag leaks, plus `thermo-nuclear-code-quality-review-subagent` for maintainability, file-size growth, spaghetti, and abstraction risks including the 1k-line rule. After both finish, thermos deduplicates overlapping findings, prioritizes evidence with file references, and returns a unified verdict. The plugin bundles 3 skills and 2 review subagents migrated from cursor-team-kit. Developers reach for thermos when a branch needs aggressive dual-pass review faster than running the full extended thermo playbook manually.
- Shorter hot-review template
- Retains severity-ranked findings
- Targets common failure classes
- Faster than full thermo pass
- Pairs with standard ship gates
Thermos by the numbers
- 602 all-time installs (skills.sh)
- Ranked #212 of 1,352 Code Review & Quality skills by installs in the Skillselion catalog
- Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/cursor/plugins --skill thermosAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 602 |
|---|---|
| repo stars | ★ 2.5k |
| Last updated | August 5, 2026 |
| Repository | cursor/plugins ↗ |
How do you run parallel security and quality PR reviews?
Invoke a condensed thermo-nuclear review workflow for medium-risk PRs that still need aggressive checking without running the full extended playbook.
Who is it for?
Developers reviewing medium-risk pull requests or feature branches who want parallel security and maintainability audits without manually coordinating two review passes.
Skip if: Developers needing a single lightweight lint pass, greenfield prototyping without a diff, or the full extended thermo playbook on every trivial change.
When should I use this skill?
The user requests thermos, double thermo review, or combined bug/security and code-quality branch audits on a PR or changed files.
What you get
Unified review report with prioritized, deduplicated bug, security, devex, and code-quality findings referencing changed files.
- unified review verdict
- prioritized deduplicated findings
By the numbers
- Launches 2 parallel thermo-nuclear review subagents per run
- Thermos plugin bundles 3 skills and 2 dedicated review subagents
- Code-quality rubric enforces a 1k-line file-size rule
Files
Thermos
Run the two thermo review passes as async background subagents in parallel, then synthesize their results.
Workflow
1. Determine the review scope from the user request, PR, current branch, or relevant changed files. 2. Gather the diff and any file/context excerpts needed for reviewers to evaluate the change without guessing. 3. Launch both subagents in the same message with run_in_background: true:
subagent_type: "thermo-nuclear-review-subagent"for bugs, breakages, security, devex regressions, feature-flag leaks, and other branch-audit risks.subagent_type: "thermo-nuclear-code-quality-review-subagent"for maintainability, structure, file-size growth, spaghetti, abstractions, and codebase-health risks.
4. Pass each subagent the same scoped diff/file context and ask it to return prioritized findings with file references and evidence. 5. After both finish, synthesize the results with findings first, deduplicated across reviewers. Weight overlapping findings more heavily, resolve disagreements with your own judgment, and keep summaries brief.
If individual background summaries are already visible to the user, do not restate them wholesale. Surface the unified verdict, the highest-signal findings, and any remaining uncertainty.
Related skills
Improve Codebase ArchitectureSafely deepen clusters of shallow modules into cohesive, testable units while respecting their external dependencies.566k196k
Caveman ReviewGet ultra-compressed, one-line code review comments that cut noise while keeping every actionable fix.270k94.7kCodebase DesignShared vocabulary for designing deep modules: improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable.266k196kCavecrewDelegate coding tasks to specialized subagents that return compressed output, keeping the main context window usable for much longer sessions.220k94.7kCode ReviewReviews a branch or PR diff on two axes at once: conformance to coding standards plus a code-smell baseline, and whether it actually implements the original spec.201k196k
Requesting Code ReviewDispatch a consistent, high-signal code reviewer subagent that catches plan deviations and quality issues before merging or continuing development.183k264kHow it compares
Pick thermos over a single thermo-nuclear review when a medium-risk branch needs both security or correctness and maintainability audits in one parallel pass.
FAQ
What does the thermos skill run in parallel?
thermos launches `thermo-nuclear-review-subagent` for bugs, security, devex regressions, and feature-flag leaks, plus `thermo-nuclear-code-quality-review-subagent` for maintainability and file-size risks. Both run as background subagents on the same scoped git diff before synthes
How is thermos different from a single thermo-nuclear review?
thermos orchestrates both deep security or correctness review and strict code-quality review in one parallel pass, then deduplicates findings. Individual `thermo-nuclear-review` or `thermo-nuclear-code-quality-review` skills cover only one rubric when a single lens is enough.
What context does thermos need before review?
thermos needs review scope from the user request, pull request, current branch, or changed files, plus the git diff and full contents of modified files. Subagents evaluate the provided diff context instead of guessing missing implementation details.