Skip to content

Spec Kitty slash commands reference — CLI quick reference

Slash Command Reference

Slash commands are invoked inside your AI agent (Claude Code, Codex CLI, Cursor, etc.). They are generated from Spec Kitty command templates and typically accept optional free-form arguments.

Syntax format in this reference:

  • COMMAND: /spec-kitty.<name> [arguments]
  • Arguments: free-form text passed as $ARGUMENTS in templates

/spec-kitty.specify

Syntax: /spec-kitty.specify [description]

Purpose: Create or update a mission specification from a natural-language description.

Prerequisites:

  • Run from the repository root checkout (no worktree).
  • If the mission should land on a branch other than the current branch, resolve that intent first with spec-kitty agent mission branch-context --json --target-branch <branch>.
  • Discovery interview is required before generating artifacts.

What it does:

  • Runs a discovery interview and confirms an intent summary.
  • Determines mission (software-dev or research).
  • Calls spec-kitty agent mission create to create mission scaffolding.

Creates/updates:

  • kitty-specs/<feature>/spec.md
  • kitty-specs/<feature>/meta.json
  • kitty-specs/<feature>/checklists/requirements.md

Optional handoff-packet intake: if spec-kitty intake staged a brief whose YAML frontmatter declares handoff_packet: 1 (an upstream requirements tool’s export), /spec-kitty.specify adopts that packet’s FR/AC ids verbatim instead of re-minting them — the discovery interview above still runs. A packet-less brief, an unrecognised version, or malformed frontmatter all degrade to today’s plain-prose intake; the command never fails because the overlay is absent or invalid. See Handoff Packet v1.

Related: /spec-kitty.plan, /spec-kitty.charter


/spec-kitty.plan

Syntax: /spec-kitty.plan [notes]

Purpose: Create the implementation plan and design artifacts based on the spec.

Prerequisites:

  • Run from the repository root checkout.
  • Spec exists for the feature.

What it does:

  • Conducts planning interrogation.
  • Calls spec-kitty agent mission setup-plan.
  • Generates planning artifacts and updates agent context files.

Creates/updates (as applicable):

  • kitty-specs/<feature>/plan.md
  • kitty-specs/<feature>/research.md
  • kitty-specs/<feature>/data-model.md
  • kitty-specs/<feature>/contracts/
  • kitty-specs/<feature>/quickstart.md
  • Agent context file (e.g., CLAUDE.md)

Related: /spec-kitty.specify, /spec-kitty.tasks, /spec-kitty.research


/spec-kitty.tasks

Syntax: /spec-kitty.tasks [notes]

Purpose: Generate work packages and task prompts from spec and plan.

Prerequisites:

  • Run from the repository root checkout.
  • spec.md and plan.md exist.

What it does:

  • Reads spec/plan (and optional research artifacts).
  • Writes tasks.md plus one prompt file per work package.
  • Calls spec-kitty agent mission finalize-tasks to populate dependencies.

Creates/updates:

  • kitty-specs/<feature>/tasks.md
  • kitty-specs/<feature>/tasks/WPxx-*.md (flat directory)

Related: /spec-kitty.plan, /spec-kitty.implement, /spec-kitty.analyze


/spec-kitty.tasks-outline

Syntax: /spec-kitty.tasks-outline [notes]

Purpose: Create the work package manifest (wps.yaml) that defines each work package’s metadata, dependencies, and file ownership. This is the first stage of staged work-package authoring; tasks.md is generated later from this manifest, not written by hand.

Prerequisites:

  • Run from the repository root checkout.
  • spec.md and plan.md exist for the mission.

What it does:

  • Reads the spec and plan to identify work package boundaries.
  • Writes wps.yaml as the single source of truth for WP id, title, dependencies, owned_files, requirement_refs, and subtasks.
  • Leaves tasks.md untouched — it is generated automatically by /spec-kitty.tasks-finalize from wps.yaml.

Creates/updates:

  • kitty-specs/<mission>/wps.yaml

Related: /spec-kitty.plan, /spec-kitty.tasks-packages, /spec-kitty.tasks


/spec-kitty.tasks-packages

Syntax: /spec-kitty.tasks-packages [notes]

Purpose: Materialize one prompt file per work package (tasks/WP*.md) from the manifest in wps.yaml. This is the second stage of staged work-package authoring.

Prerequisites:

  • Run from the repository root checkout.
  • wps.yaml exists with complete WP definitions (written by /spec-kitty.tasks-outline).

What it does:

  • Reads wps.yaml and generates an independent prompt file for each work package.
  • Updates wps.yaml with per-WP owned_files, requirement_refs, subtasks, and prompt_file values, and assigns an agent profile per WP.

Creates/updates:

  • kitty-specs/<mission>/tasks/WP*.md
  • kitty-specs/<mission>/wps.yaml (per-WP details)

Related: /spec-kitty.tasks-outline, /spec-kitty.tasks-finalize


/spec-kitty.tasks-finalize

Syntax: /spec-kitty.tasks-finalize [options]

Purpose: Finalize the mission’s work packages — validate dependencies and requirement mapping, generate tasks.md, update WP frontmatter, and commit the task artifacts to the target branch. This is the final stage of staged work-package authoring.

Prerequisites:

  • Run from the repository root checkout.
  • WP prompt files and wps.yaml exist (from /spec-kitty.tasks-packages).

What it does:

  • Runs spec-kitty agent mission finalize-tasks, starting with a --validate-only preflight.
  • Validates dependencies (no cycles, no invalid references) and requirement mapping against spec.md.
  • Computes lanes, generates tasks.md, updates WP frontmatter, and commits the artifacts automatically.

Creates/updates:

  • kitty-specs/<mission>/tasks.md
  • kitty-specs/<mission>/tasks/WP*.md (frontmatter: dependencies, requirement refs, lanes)

Related: /spec-kitty.tasks-packages, /spec-kitty.implement, /spec-kitty.analyze


/spec-kitty.implement

Syntax: /spec-kitty.implement [WP_ID]

Purpose: Resolve the execution workspace and start implementation for a specific work package.

Prerequisites:

  • Work packages exist in kitty-specs/<feature>/tasks/.
  • Run from the repository root checkout for the action prompt; the execution workspace is created or reused by the CLI.

What it does:

  • If explicit slash-command args are provided, forwards the WP selection into the resolver-first action flow.
  • Step 1: spec-kitty agent action implement WP## --agent <agent> to show the prompt and move the WP to doing.
  • Step 2: spec-kitty implement WP## to create or reuse the execution workspace.
  • Implementation happens inside the resolved workspace path printed by the command.

Creates/updates:

  • .worktrees/<feature>-lane-<id>/
  • kitty-specs/<feature>/tasks/WP##-*.md lane status updates

Related: /spec-kitty.tasks, /spec-kitty.review


/spec-kitty.review

Syntax: /spec-kitty.review [WP_ID]

Purpose: Review a completed work package and update its lane status.

Prerequisites:

  • Run from any checkout where the mission can be resolved; review will attach to the canonical execution workspace if needed.
  • WP must be in lane: "for_review".

What it does:

  • If WP_ID is provided, forwards it to the resolver-first workflow as an explicit --wp-id.
  • Loads the WP prompt, supporting artifacts, and code changes.
  • Performs structured review and records feedback via --review-feedback-file.
  • Persists feedback in shared git common-dir and writes frontmatter review_feedback pointer (feedback://...) in the WP file.
  • Moves the WP to approved (review passed, merge pending) or back to planned (needs changes). The merge workflow later records done.
  • Updates tasks.md status when approved.

Creates/updates:

  • kitty-specs/<feature>/tasks/WP##-*.md (review feedback, lane changes)
  • kitty-specs/<feature>/tasks.md (checkbox status)

Related: /spec-kitty.implement, /spec-kitty.accept


/spec-kitty.accept

Syntax: /spec-kitty.accept [options]

Purpose: Validate mission readiness and generate acceptance results.

Prerequisites:

  • Run from any checkout or branch where mission auto-detection works.
  • All WPs should be approved or done, with review feedback resolved.
  • Run this after the implement-review loop and before /spec-kitty.consolidate.

What it does:

  • Auto-detects mission slug and validation commands when possible.
  • Runs spec-kitty agent mission accept to perform acceptance checks.
  • Outputs acceptance summary and merge instructions.

Creates/updates:

  • Acceptance output in the mission directory (and optional commits depending on mode)

Related: /spec-kitty.review, /spec-kitty.consolidate


/spec-kitty.consolidate

Syntax: /spec-kitty.consolidate [options]

Purpose: Consolidate an accepted mission into the target branch and clean up worktrees.

Prerequisites:

  • Run from any checkout where the mission can be resolved (repository root checkout or execution workspace).
  • By default, consolidation lands in the mission’s recorded target branch; use --target <branch> only when you intentionally want to override it.
  • Mission must pass /spec-kitty.accept.

What it does:

  • Executes spec-kitty consolidate with selected strategy and cleanup flags.
  • Optionally pushes to origin and deletes worktrees/branches.
  • After merge, run /spec-kitty-mission-review, then surface the retrospective captured at the runtime terminus. Canonical post-merge sequence (FR-019):
    1. Mission review — /spec-kitty-mission-review confirms spec→code fidelity.
    2. Author or verify the retrospective — under default policy, the runtime authored it during merge; verify via cat .kittify/missions/<mission_id>/retrospective.yaml. If absent (older mission, generator failure under warn policy), author with spec-kitty retrospect create --mission <slug>.
    3. Surface findings — spec-kitty retrospect summary aggregates across missions (read-only); spec-kitty agent retrospect synthesize --mission <slug> previews or applies proposals from one record (dry-run by default; add --apply to mutate).

Creates/updates:

  • Merges mission branch into target branch.
  • Deletes worktree and/or mission branch depending on flags.

Related: /spec-kitty.accept


/spec-kitty.status

Syntax: /spec-kitty.status

Purpose: Display current kanban status for work packages.

Prerequisites:

  • Run from a repo or worktree with access to kitty-specs/<feature>/tasks/.

What it does:

  • Runs spec-kitty agent tasks status.
  • Shows a lane-based status board, progress metrics, and next steps.

Creates/updates: None (read-only).

Related: /spec-kitty.tasks, /spec-kitty.implement


/spec-kitty.charter

Syntax: /spec-kitty.charter

Purpose: Create or update the project charter.

Prerequisites:

  • Run from the repository root checkout.

What it does:

  • Runs a phase-based discovery interview (minimal or comprehensive).
  • Writes project-wide principles to the charter file.

Creates/updates:

  • .kittify/charter/charter.md

Related: /spec-kitty.specify, /spec-kitty.plan


/spec-kitty.research

Purpose: Capture discovery findings before a spec or plan exists.

Prerequisites:

  • Run from any checkout where the mission can be resolved. No spec or plan is required.

What it does:

  • Tells you to record decisions, evidence, and open questions directly in research.md (create or extend it; never truncate existing content).

For mission types that ship research templates (today: the research mission type), once that mission’s plan is filled in you can additionally scaffold research.md, data-model.md, and the two CSV stubs from the mission type’s shipped templates:

Syntax: spec-kitty research --mission <handle> [--force]

Creates/updates (research-type missions only — creates nothing for mission types without research templates, e.g. software-dev):

  • <mission_dir>/research.md
  • <mission_dir>/data-model.md
  • <mission_dir>/research/evidence-log.csv
  • <mission_dir>/research/source-register.csv

Related: /spec-kitty.specify, /spec-kitty.plan


/spec-kitty.analyze

Syntax: /spec-kitty.analyze [notes]

Purpose: Cross-artifact consistency analysis after tasks generation.

Prerequisites:

  • Run from any checkout where the mission can be resolved.
  • spec.md, plan.md, and tasks.md must exist.

What it does:

  • Reads spec, plan, tasks, and charter (if present).
  • Produces and persists an analysis report of gaps and conflicts.
  • Records source artifact hashes so implementation can reject stale analysis.

Creates/updates:

  • kitty-specs/<feature>/analysis-report.md

Related: /spec-kitty.tasks, /spec-kitty.implement

Getting Started

Practical Usage