kagura-ai

Audit .claude/ configuration against current codebase

Upload, list, delete, and get download URLs for files in Kagura Memory's R2 storage (sha256 integrity-bound; an upload can optionally be bound to an owning context for access control). Use when the user wants to attach or manage raw files in their workspace.

Store and retrieve zero-knowledge secrets with the `kagura secret` CLI — age recipient encryption, local decrypt (memory-cloud never sees plaintext). Use for keygen/enroll, put/get, grant/revoke/rotate, owner-only delete, exec-with-injected-env, and audit verification.

Manage Kagura Memory authentication — log in (OAuth device flow), check status, refresh tokens, and list or switch profiles. Use when the user wants to sign in to Kagura, see which account/workspace is active, or manage multiple profiles.

Diagnose the local Kagura Memory CLI setup (install, auth, MCP wiring). Use before other kagura skills, or whenever a kagura command reports a blocked or misconfigured environment.

Ingest a document or URL into Kagura Memory — extract text, chunk it, and store an overview plus per-section memories. Use when the user wants to add a PDF, Office/HTML/EPUB file, audio/video, or a web page to their memory graph.

Manage Kagura resource tokens and push external data (Slack/CI/CRM events) into a context. Use for resource token create/list/update/revoke, resource setup/import, and impact/schema stats.

Wire Claude Code to Kagura Memory Cloud by writing the MCP config. Use during onboarding to connect this machine's Claude Code to a Kagura workspace.

Use when kagura-engineer:doctor reports a blocked or failing environment — scaffolds missing config, installs and authenticates harness prerequisites, and generates Memory Cloud MCP config by shelling out to `kagura-engineer setup`, then re-checks with doctor.

Use before any kagura-engineer run/review, or when the harness reports a blocked environment — diagnoses the local setup (git, selected brain CLI, gh, ollama, memory backend, and gh-issue-driven plugin) by shelling out to `kagura-engineer doctor` and reports what must be fixed.

Use to measure whether memory grounding actually improves PR quality — shells out to `kagura-engineer eval` with one or more issue numbers, running the SAME fixed issue set in two arms (recall ON vs OFF) and printing an A/B table. HARNESS — high cost; it runs the full loop twice per issue and mutates the repo. Confirm with the user before launching.

Use to review a git ref, branch, or PR with the local kagura-engineer reviewer — shells out to `kagura-engineer review [target]`, returning a structured verdict and findings. With `--fix` it is a HARNESS that edits files and commits; without `--fix` it is read-only.

Use to drive a single GitHub issue to a pull request with the kagura-engineer harness — shells out to `kagura-engineer run` with an issue number (guard → recall → worktree → start → implement → ship → persist). HARNESS — this mutates the repo, creates a PR, and spends model budget; confirm with the user before launching.

Use to bootstrap a fresh checkout for the kagura-engineer harness — scaffolds a commented repo.yaml template and adds it to .gitignore by shelling out to `kagura-engineer init`. Run this before setup/run/review when no repo.yaml exists yet. Idempotent and never overwrites.

Use Kagura Memory Cloud from Codex for long-term project memory. Trigger when the user asks to start or restore a session, recall past decisions or bug fixes, save session knowledge, remember a decision or troubleshooting note, inspect memory contexts, or verify the kagura-memory MCP connection.

Use when the user wants a memory-grounded implementation plan or decision record for an idea — recalls related past decisions and known traps, produces a recallable plan doc, and remembers it. Do NOT use to build or execute the plan (that is the engineer's job).