heracles

Plugin heracles from marketplace heracles-skills in davidtobonm/heracles-skills. Commands read from the repo's manifest — 19 days ago.

Skills (26)

• arena

  Spawn N parallel candidates at the same task, pick a base, graft the strongest parts of the losers into it. Called by design-architecture for one-way decisions, or when the user types /arena or says 'arena this'.

• build-task

  Build one planned sprint task end to end on its own branch. Loads the task and its stories, isolates the work in a worktree, builds test-first (or settles a design decision with the owners), proves each acceptance criterion with evidence, runs the checks, commits and reports. Use for /build-task <sp

• close-sprint

  Close a sprint after its release is merged into main and the humans call it done. Brings the living docs (product, glossary, architecture, contracts, diagrams) back in line with the code, moves leftovers to the backlog, runs a short retro that turns repeated mistakes into checks, and freezes the spr

• codebase-design

  Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

• create-verification-skill

  Generate a project-local verification skill that drives your app the way a user does — any language, framework, or platform. Use for /create-verification-skill, \"make a control skill for this repo\", or when a project has no scripted way to prove UI/CLI/service behavior.

• define-product

  Define or change what the product is. Interviews the owners and writes the product definition, the glossary, the user story map with story ids, and the ordered backlog. Use for /define-product, at the start of a project, or when a product rule, a user role or the scope of an epic changes.

• design-architecture

  Design or change the system architecture (runtimes, services, jobs, stores, data model, cloud services, deploy) and the module shape inside each service. Settles each choice with the owners, runs competing candidates for one-way choices, records decisions, and writes the architecture overview, the L

• diagnosing-bugs

  Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

• domain-modeling

  Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing the glossary in docs/product/glossary.md, or deciding whether a decision needs an ADR.

• draw-diagram

  Create or update a diagram in the repo docs (system architecture, data flow, domain model, storage schema) as diffable text that renders on GitHub, and check that it parses. Use when a design skill settles a structure, when a change makes a diagram stale, or when asked to draw or update a diagram.

• grilling

  Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.

• handoff

  Compact the current conversation into a handoff document for another agent to pick up.

• init

  Set up this repo for the heracles workflow. Adds AGENTS.md, the docs skeleton, the check scripts, the subagent role files and, per agent the team uses, CLAUDE.md, Claude Code settings and hooks, Cursor hooks, GEMINI.md or OpenCode subagents, without overwriting any existing file. Use once per repo,

• maintain-verification-skill

  Periodic pass that keeps a project's verification skill and feature map honest: parallel source readers per feature, one live session driving every feature, at most one PR of proven corrections. Use for /maintain-verification-skill or \"audit the verify skill\".

• plan-sprint

  Plan the next delivery sprint. Takes stories from the story map, interviews the owners until each has checkable acceptance criteria, has to-spec write the sprint spec and to-tickets write bounded tasks, groups them in waves (with scope, seams, verify command, model and effort per task), has a separa

• record-decision

  Write, update or supersede an Architecture Decision Record (ADR) in docs/decisions/. Use when a decision is hard to reverse, surprising without context and the result of a real trade-off: when a design, planning or build skill settles one, and also when any conversation lands on one unplanned, for e

• review-change

  Review a branch's changes against its sprint spec and this repo's rules with independent reviewer subagents, give a 1-5 merge score, and optionally fix and re-review in a bounded loop. Use for /review-change, before opening or merging a PR, when asked to "review this", or when ship-change calls it.

• run-sprint

  Build a planned sprint as its orchestrator. Works from dev, runs each wave's tasks in parallel implementer subagents on task branches in their own worktrees, audits every task's diff and evidence, opens a PR into dev for each passing task and merges it, keeps the plan's progress, and stops for the o

• ship-change

  Take a finished branch to a merge-ready pull request: a task or design branch into dev, or dev into main as a release. Rebases, reruns the checks, runs the review loop, writes the PR from the sprint spec and evidence, opens it, and watches CI until green. The main session may merge PRs into dev when

• skill-guide

  Tell the human which skills and commands fit what they want to do, in order, from the heracles plugin and the skills in this repo. Starts nothing. Use for /skill-guide <what you want to do>, or when someone asks "which skill should I use", "how do I do X here" or "what skills do we have".

• tdd

  Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

• to-spec

  Write a sprint spec (problem, user stories, checkable acceptance criteria, test seams, out of scope) into a given spec.md from what the conversation has already settled. No interview. Called by plan-sprint after its interview, or when asked to turn a discussion into a spec.

• to-tickets

  Break a sprint spec into tracer-bullet tasks (vertical slices, each with its blocking edges and a verify command), quiz the humans on the breakdown, and write the approved tasks into a given plan.md. Called by plan-sprint after to-spec, or when asked to break a spec into tasks.

• unslop

  Cut AI tells from writing that will be published, such as a PR title and body, a doc or an ADR. Called by ship-change, or when the user asks to unslop a text.

• verify-sprint

  Prove a built sprint does what its spec promised. A fresh verifier checks every acceptance criterion against the running app, in a seeded demo or test environment, and the code, then the owners accept each user-facing criterion one at a time. Failures become fix tasks. Use for /verify-sprint <sprint

• whats-next

  Say where the project stands and which command to run next, from docs/status.md, the active sprint plan and the backlog. Use for /whats-next, when a session starts, after a pause, or when someone asks "what's next" or "where are we".

Similar plugins