Skills for running long unattended AI coding sessions. It will help the agent verify and review its own work using tests, adversarial review, and manual verification through cURL or real browsers. This skill reduces the burden on human review and feedback by making the agent more thorough and self-sufficient.
Skill Claude CodeCodex
Part of backpressured
Sets up backpressure for a project — detects its check/run commands, writes or edits a BACKPRESSURE.md, verifies Playwright, and installs the MCP servers the loop uses. Runs ONLY when the user explicitly invokes /backpressured:config; never auto-trigger it.
Skill Claude CodeCodex
Part of backpressured
Use when a backpressured code-review subagent is reviewing a diff for correctness/logic, simplicity/reuse, or test-quality defects — everything except type design. The default reviewer dimension.
Skill Claude CodeCodex
Part of backpressured
Runs a backpressured automated development loop. Use it ONLY when the user explicitly mentions "backpressured" or "backpressure" — e.g. asking for a "backpressured" loop/run, "do this backpressured", or invoking it by that name — or runs the /backpressured:goal command. A bare /goal, or any goal-like request that does…
Skill Claude CodeCodex
Part of backpressured
Verifies a feature works by running the real app, not by trusting green tests. Use when a backpressured loop reaches the before-done gate (Phase 3) and needs to confirm new behavior end-to-end — curling live API endpoints and driving a real browser via Playwright — including the cheap unhappy paths, before the task is…
Skill Claude CodeCodex
Part of backpressured
Reviews implementation plans during iterative backpressured development. Use when a backpressured loop's Phase 1 reviewer subagent is judging whether a lightweight implementation plan's approach/architecture is sound before any code is written — approving it or sending it back with approach-level concerns.
Skill Claude CodeCodex
Part of backpressured
Use after a pull request has been opened and the task is not finished until it lands. Triggers on "PR is open", "CI is running", "is it merged yet", a PR that is green but not merged, fresh CI that might still flip, or being tempted to report a task done at PR-open. GitHub / gh CLI based.
Skill Claude CodeCodex
Part of backpressured
Use when a backpressured loop needs to run benchmarks on a performance-sensitive project and decide whether a change is a regression, an improvement, or a wash — per-iteration sanity checks and the full pre-done run.
Skill Claude CodeCodex
Part of backpressured
Use when a backpressured code-review subagent is judging whether a diff uses the type system to make illegal states unrepresentable — i.e. the diff touches data models, function signatures, or domain types. TypeScript-centric, with mappings to Rust, Swift, Kotlin, and Haskell/OCaml.
Skill Claude CodeCodex
Part of backpressured
Use when a backpressured loop has a front-end/UI change and needs to compare the rendered result against its design reference (a Figma frame or images on a Linear ticket) — catching layout, spacing, contrast, and visual-consistency defects automated tests and happy-path clicking miss.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: