Describe a change and ac builds it: parallel research, slices routed to haiku, sonnet or opus by cost, an independent review, and an unattended mode that ends at a read-only gate. Adds a planner, a roadmap coordinator, a QA pass, authoring skills and gated, fail-open hooks.
★not rated 3
changed 6d agoA
tokens not measured
originalMIT
Give every git worktree its own isolated, disposable environment (app port, ephemeral Chrome/CDP, per-worktree database) behind a single .env.worktree contract, so N agentic coding sessions build features in parallel without colliding on ports, the browser, or the database.
★not rated 3
changed 24d agoA
tokens not measured
originalMIT
Plays a repeating sound when Claude Code is genuinely blocked waiting on you (a permission prompt, an explicit needs-input signal, or a dialog) and stops automatically the moment you respond. Interval and sound are configurable.
★not rated 3 2mo agoA
tokens not measured
originalMIT
Time awareness for coding agents, in Claude Code and Codex: in Claude Code, stamps every message with the local time and how long the turn took; in both, tells the agent when a turn runs long, a tool call is slow or work resumes after a gap, remembers how long each command usually takes in a project, and reports where.
★not rated 3
changed 1mo agoA
tokens not measured
originalMIT
Automated hook-driven test framework generator for Claude Code projects. Scaffolds complete test infrastructure including commands, hooks, runner scripts, and test definitions into target projects.
Use Cursor's CLI (cursor-agent) from Claude Code to delegate coding tasks and review code, choosing the model per call. Shells out to your locally installed, authenticated cursor-agent (no data is sent to the plugin author). Registers session-lifecycle hooks and an optional, off-by-default stop-time review gate.
★not rated 3 4mo agoA
tokens not measured
originalMIT
Generate ADE-Bench benchmark tasks from your own dbt project. Scans your models, proposes realistic bug-injection scenarios, and writes the task scaffolding (config, patches, scripts, custom assertion tests) ready to run against AI agents.
★not rated 3 6mo agoA
tokens not measured
originalMIT
Self-contained session state preservation for Claude Code context compaction: state capture, recovery injection, resume recovery, and built-in context usage warnings.
★not rated 3 3mo agoA
tokens not measured
originalMIT
macOS screen recorder with system audio via ScreenCaptureKit, CLI, no driver.
★not rated 3 2mo agoA
tokens not measured
originalMIT
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: