Feature Index

GENERATED — do not edit. Run scripts/cct list --write (or scripts/generate-feature-index.sh) to regenerate. Sources of truth: shared/features/catalog.yaml, shared/skills/, adapters/claude-code/.claude/commands/, and shared/capabilities/catalog.yaml. To change an entry, edit the source — a drift guard (--check) fails the build if this file is stale.

19 features · 14 slash commands · 24 skills · 20 capabilities

New here? Start with the Quick Start, then browse the tables below. In a session, run scripts/cct list to print this on demand.

Features

What the harness offers, one row per user-facing feature. The title links to the primary guide. Maturity and release state are defined in maturity.md.

Feature	What you get	Maturity	Since
Spec-driven development	A feature is specified, planned and task-listed in specs/<feature-id>/ before any code is written, with an origin-alignment gate that stops the build when the working spec drifts from what the user asked for.	stable	v1.1.0
Four-phase agent workflow	Work moves through Research, Plan, Build and Review with a single agent in every phase but Build, where a Team Lead delegates to sub-agents; phase transitions are gated on verification.	stable	v1.1.0
Shape-Up product bets	Rough ideas become pitches with an appetite, scopes and a circuit breaker; bets run against a hill chart and end in a retrospective and a cooldown report.	stable	v1.1.0
Peer review by a second model	A second provider reviews the diff in a read-only sandbox and returns a PASS, FAIL or INCONCLUSIVE verdict with findings; the phase cannot complete on a FAIL, and a failed reviewer falls back down a chain.	stable	v1.1.0
Unattended auto-build runs	An approved feature is built phase by phase by a driver outside the session, with a gating reviewer per phase, bounded fix sessions, a cost and wall-clock cap, a ledger, and a PR at the end.	beta	unreleased
Cost and safety caps	An unattended run stops on its own when spend, wall-clock, review rounds or session turns exceed the caps frozen at admission, and the Studio raises an alert at 80% and 100%.	beta	unreleased
Coverage contract	A run’s verification.yaml binds each requirement to the test that proves it, and the driver refuses a phase whose coverage fell below the frozen contract.	beta	unreleased
Runtime conformance evaluator	The built application is started and probed against the contract’s runtime criteria before the phase can land, so a green unit suite cannot hide a service that does not come up.	experimental	unreleased
Visual verification gate	A UI feature’s run captures screenshots against the contract’s visual criteria and parks rather than passes when the capture cannot be made.	experimental	unreleased
Ralph Loop single-agent iteration	One agent iterates on a task until the verification command passes or the iteration cap is reached, with no delegation.	stable	v1.1.0
Project templates and sync	A new project starts from a stack-specific CLAUDE.md with an agent team, slash commands and CI workflows, and an existing project can be re-synced to the latest template.	stable	v1.1.0
UI design harness	Generated UI is steered by a committed DESIGN.md and design tokens and gated by a visual-review loop with an accessibility check and an anti-slop rubric.	stable	v1.1.0
LLM wiki maintainer	Lessons and decisions are promoted into a curated project wiki that agents consult before re-reading raw sources, with an ingest pipeline and a lint.	stable	v1.1.0
Benchmark harness	Coding agents are run against a task set through pluggable backends and scored with a judge, so a configuration change is measured rather than felt.	beta	v1.1.0
Session analytics	Claude Code, Pi and Aider sessions are ingested into a store and queried for cost, turns, tool use, phases, labels and struggle signals from a CLI or an MCP server.	stable	v1.1.0
Studio web UI	The analytics, benchmarks, auto-build runs, team alerts and the project’s own documentation are browsed in one local web app with a settings page for the judge and embedding providers.	beta	unreleased
LLM routing profiles	Auto-build tasks are routed to a backend by route class and role under an execution profile, with validate, status and explain commands and a scheduled tick.	experimental	unreleased
MemKernel persistent memory	Session context is recalled at start, checkpointed before compaction and recovered after it by hooks that activate only when MemKernel is installed.	experimental	v1.1.0
cct command-line front door	One provider-neutral command lists every slash command, skill and capability and drives routing; doctor and config are planned.	experimental	unreleased

Adapter support

How each adapter delivers the feature (defined in maturity.md): enforced means a runtime gate in that adapter can block, advisory that it receives the feature as content it cannot enforce, — unsupported, and n/a that the feature runs outside any adapter.

Feature	aider	claude-code	codex	cursor	github-copilot	pi	windsurf
Spec-driven development	advisory	enforced	advisory	advisory	advisory	enforced	advisory
Four-phase agent workflow	advisory	enforced	advisory	advisory	advisory	enforced	advisory
Shape-Up product bets	advisory	enforced	advisory	advisory	advisory	advisory	advisory
Peer review by a second model	advisory	enforced	advisory	advisory	advisory	advisory	advisory
Unattended auto-build runs	—	enforced	enforced	—	—	enforced	—
Cost and safety caps	—	enforced	enforced	—	—	enforced	—
Coverage contract	—	enforced	enforced	—	—	enforced	—
Runtime conformance evaluator	—	enforced	enforced	—	—	enforced	—
Visual verification gate	—	enforced	enforced	—	—	enforced	—
Ralph Loop single-agent iteration	advisory	enforced	advisory	advisory	advisory	advisory	advisory
Project templates and sync	advisory	enforced	advisory	advisory	advisory	advisory	advisory
UI design harness	advisory	enforced	advisory	advisory	advisory	advisory	advisory
LLM wiki maintainer	advisory	enforced	advisory	advisory	advisory	advisory	advisory
Benchmark harness	n/a	n/a	n/a	n/a	n/a	n/a	n/a
Session analytics	n/a	n/a	—	—	—	n/a	—
Studio web UI	n/a	n/a	n/a	n/a	n/a	n/a	n/a
LLM routing profiles	—	enforced	enforced	—	—	enforced	—
MemKernel persistent memory	advisory	enforced	advisory	advisory	advisory	—	advisory
cct command-line front door	n/a	n/a	n/a	n/a	n/a	n/a	n/a

Slash commands

Type these in an agent session (e.g. Claude Code).

Command	What it does
/auto-build	Scaffold and explain an autonomous build run for an approved SDD feature. Validates the approval + origin gates, writes specs/<feature-id>/automation.json from the template, and prints the driver command — the driver itself runs OUTSIDE this session.
/bet	Lock a shaped Shape-Up pitch as a bet for the next cycle.
/cooldown	Close out a cycle — invoke cooldown-report and finalize the active pitch’s bet status.
/cycle-start	Begin a cycle for a bet pitch — initializes the hill chart and transitions the pitch to building.
/hill	Update a scope’s status on the hill chart of the active pitch.
/list-agents	List the custom subagents installed on this machine and in this project — the replacement for the built-in /agents command removed from recent Claude Code versions.
/origin-check	Run the origin-confirmation circuit breaker for a feature. Validates that the working spec/plan is a faithful realisation of the user’s original idea, and surfaces an interactive escalation if it has drifted.
/phase-complete	Signal phase completion. Validates that the review loop has passed (if peer review is enabled) and runs the post-phase checklist.
/promote-lesson	Promote a session-level lesson into the project wiki at knowledge/wiki/. Follows the canonical curator procedure; never edits the wiki without it.
/ralph-start	Start a Ralph Loop for autonomous task completion.
/retro	Run a structured retrospective for the current session. Proposes rule and documentation changes but does not apply them.
/review-decide	Resolve a circuit breaker in the review loop. Accepts exactly one argument: approve, reject, or retry.
/review-submit	Submit work for peer review. Validates clean worktree, initializes or continues the review loop, invokes review-round-runner.sh, and returns the verdict.
/shape	Shape a rough idea into a Shape-Up pitch via the pitch-shaper agent.

Skills

On-demand instruction modules the agent loads by phase or when relevant.

Skill	Description
agent-team-protocol	Multi-agent delegation rules, three-phase workflow (Plan/Build/Review), model selection, collaboration gates, and Ralph Loop integration.
auto-build-loop	Autonomous build driver after SDD spec approval: phase-scoped headless build sessions, driver-owned commits, cross-provider review gates, milestone pauses, and fail-closed escalation — under explicit opt-in autonomy profiles.
clarification-protocol	When and how to ask clarifying questions before implementing. Data model review gate and ambiguity resolution rules.
coding-standards	Quality gates, prohibited patterns, verification discipline, and self-audit rules for all code generation and review sessions.
copilot-conventions	Cross-copilot portable conventions: target alignment, complexity control, minimal changes, git discipline, and project structure.
copyright-headers	Copyright header rules for generated source files. Applies the company name from providers.toml to new files.
design-system	Derive a unique, domain-fit design direction and enforce design-token + anti-slop discipline. Read/author DESIGN.md before building any UI; override framework defaults so output is bespoke by construction, not generic.
environment-setup	Environment variable patterns, config file validation, and setup verification for new project scaffolding.
infra-verification	Infrastructure artifact verification: Docker builds, CI workflows, and launcher flags must be executed, not just syntax-checked.
integration-testing	Test integration points early. Verify cross-service contracts, API boundaries, and data flow before declaring done.
memkernel-memory	MemKernel persistent memory protocol. Self-guarding: activates only when the memkernel MCP server is configured.
opus-4-7-features	Optional guidance for sessions using Claude Opus 4.7: adaptive thinking, xhigh effort, prompt caching, /btw side questions.
origin-confirmation	Origin-confirmation circuit breaker: machine-checkable origin frontmatter, alignment-check protocol, three gates (plan-approval, build-entry, phase-complete), and interactive escalation on deviation.
phase-workflow	Phase transition rules, post-phase verification steps, peer review validation, and commit-gate approval flow.
provider-collaboration-protocol	Cross-provider peer review protocol: session flags, review flow, collaboration artifacts, fail-closed enforcement, and CI validation.
ralph-loop	Single-agent autonomous iteration loop: PRD-driven, self-verifying, with configurable iteration limits and completion criteria.
review-loop	Agent-driven peer review loop: structured findings, dispositions, circuit breakers, commit strategies, sandbox isolation, and plan-phase advisory mode.
safety	Non-negotiable safety constraints: destructive action guards, blocked operations, secrets policy, password storage, input validation, and dependency review.
spec-workflow	SDD specification protocol: risk-based spec_mode classification, required sections per mode, plan approval gate, and artifact directory conventions.
stack-constraints	Stack version pinning and dependency compatibility guards. Prevents silent version drift across project boundaries.
team-lead-efficiency	Build team lead efficiency rules: limit sub-agents, polling discipline, no redundant re-work, parallel task launching.
token-efficiency	Token economy rules: diff-over-rewrite, context compression, avoid verbose re-explanations, minimize round-trips.
visual-review	Closed visual-review loop for generated UI: render the running app, screenshot per breakpoint, run an axe-core a11y gate, critique against the DESIGN.md rubric, triage findings, and iterate to a quality bar with a hard cap. Drives the shippable harness/ runner.
wiki-first-query	Wiki-first query convention: consult knowledge/wiki/index.md and the linked pages BEFORE re-reading raw sources for a project topic. The wiki is the canonical project memory layer.

Capabilities

Tool-agnostic capabilities resolved per adapter. Full compatibility matrix: shared/capabilities/COMPATIBILITY.md.

Capability	Description	Default
skills.shared	Shared CCT skills are discoverable by the agent as first-class instruction resources.	disabled
prompts.commands	CCT slash commands are available as prompt templates or commands.	disabled
config.layered	Configuration resolves through an explicit layer precedence, and every resolved value can be explained with its provenance.	disabled
config.trust-gating	Project-supplied configuration is loaded only after positive trust resolution; unknown trust fails closed.	disabled
workflow.sdd	Risk-scaled SDD artifacts gate protected work; unresolved clarification markers block it.	disabled
workflow.phases	A persistent Research -> Plan -> Build -> Review phase machine governs what the agent may do in each phase.	disabled
permissions.engine	Tool, command, and path requests resolve to allow / ask / deny, with a deterministic resolution for headless sessions.	disabled
permissions.protected-paths	Sensitive paths are protected against write and traversal, with symlink and canonicalization defenses.	disabled
providers.pi	Pi can act as a peer-review provider for cross-provider review, under a read-only reviewer contract.	disabled
review.enforcement	Mandatory peer review is enforced before a build/review phase can complete, via the provider-neutral review-loop runner.	disabled
verification.enforcement	Verification gates (tests, drift, docs, …) block phase completion when required; gate types with no runnable substrate are reported, not faked.	disabled
integrations.mcp	Model Context Protocol servers can be attached, with their provenance and permissions reported.	disabled
integrations.hosted-platform	Anthropic-hosted platform services. CCT does not implement these; they are reported as external-platform so parity is never overclaimed.	unavailable
memory.session-state	Session state (workflow phase, active feature) persists across restarts and is recovered into context at session start. Pre-compaction checkpoint when the harness exposes a compaction event; otherwise checkpoints are taken at explicit CCT actions.	disabled
security.sandbox	Sandbox detection (host-unrestricted / permission-gated-only / containerized / micro-vm / remote-sandboxed) and the autonomous/ci rejection of unrestricted host execution absent an explicit override. Permissions are NOT sandboxing (reported separately, spec P5).	disabled
security.env-scrub	Name-based scrubbing of credential-shaped environment variables at harness-controlled spawn boundaries (subagent child sessions, worker handoffs), with a trust-asymmetric configuration surface: in-repo configuration may tighten but never loosen the policy. Values are never read or logged; the primary interactive session’s environment is out of scope by design.	disabled
memory.promotion	Explicit promotion/deletion of durable memories with provenance, wiki-first retrieval, and sensitive-memory controls (a fact carrying a secret is refused). Optional MemKernel (MCP) backend for code-aware retrieval.	disabled
agents.subagents	Named subagents with separate context and per-agent model/thinking/tools/permissions/skills, expressed as neutral manifests; Claude .claude/agents files import into that schema. Live child-session execution (spawn, result contracts, caps, cancellation) is a separate concern gated on the runner surface.	disabled
agents.worktrees	Parallel workers isolated in git worktrees, with worker / branch / worktree / task / owned-area / verification / merge / cleanup state tracked. Verification and merge EXECUTION are separate concerns; this capability is the isolation + tracked state + safe lifecycle.	disabled
agents.teams	Opt-in agent teams: lead/teammate identities, a shared task ledger with assignment/claiming, peer messaging, plan approval, bounded concurrency, controlled shutdown, result synthesis and partial-failure handling. DISTINCT from subagent delegation (peers coordinating over shared state, not a parent spawning children).	disabled