Canhada-Labs
OrganizationGovernance & auditability layer for Claude Code: run a structured team of specialist agents under a CEO protocol — plan→debate→execute gating, tamper-evident audit log, cross-LLM pair-rail.
Categories
Indexed Skills (48)
agent-architect
Meta-agent that drafts a new squad bundle (team-personas, pitfalls, skill-selection, personas roster, rationale) from an Owner-supplied domain brief. Read-only on canonical paths; writes ONLY into the sandboxed plan subdir. Sentinel-gated mutations.
ai-llm-orchestration
AI system and LLM Council management for {{PROJECT_NAME}}. Covers the 3-model council architecture (Claude + Gemini + GPT), snapshot construction and minimization, prompt engineering for market analysis, usage throttling (100K tokens/day budget), input injection prevention, proprietary data protection, signal quality assessment, confidence thresholds, fallback chains, and cost optimization. Use when working on any code in src/ai/, reviewing AI prompts, implementing new AI features, debugging AI signal quality, managing LLM costs, or reviewing AI security (injection attacks, data leakage). 18 files, ~9.3K lines. Also use when the user mentions "AI council", "LLM", "market analysis AI", "AI signals", or "prediction confidence".
architecture-decisions
Architecture decision-making framework for {{PROJECT_NAME}}. Covers ADR (Architecture Decision Record) format, trade-off analysis matrices, dependency graph analysis, blast radius assessment, system boundary identification, scalability testing ("10x scale" rule), when to refactor vs rewrite, multi-process architecture patterns, and cross-cutting concern management. Use when evaluating changes that touch 3+ modules, planning new features, assessing refactoring proposals, reviewing system boundaries, or making decisions that will be hard to reverse. This is the VP Engineering archetype's operating manual for architectural governance.
chaos-and-resilience
Chaos engineering, resilience patterns, failure recovery, and fault tolerance for the {{PROJECT_NAME}}. Covers circuit breaker patterns (HTTP, WS, per-venue), reconnect gate design, graceful shutdown protocol, backpressure strategies (PubSub 1MB/4MB, persistence queues), worker lifecycle management, IPC channel health detection, failure scenario matrix, bounded queue design, and event loop stall recovery. Use when reviewing or writing any code that touches error handling, retry logic, reconnection, circuit breakers, worker management, IPC channels, queue management, backpressure, graceful shutdown, health checks, or any failure-adjacent code path.
core-code-intelligence-lsp
Engineering doctrine for using Language Server Protocol tools in agent code-analysis workflows for {{PROJECT_NAME}}. Covers per-language LSP server selection, diagnostic triage, type-query patterns, and the critical discipline of anchoring analysis to structured semantic data rather than string matching. Use when performing non-trivial code review, refactoring, type-safety analysis, or any task where grep alone cannot distinguish a definition from a reference or a type from its alias.
code-review-checklist
Structured code review process for the {{PROJECT_NAME}}. Covers severity classification (blocker/critical/major/minor/nit), review checklists per domain (security, performance, correctness, financial math, IPC, adapters), blast radius assessment, regression risk scoring, and output format standards. Use when reviewing any code change before merge, evaluating PR quality, assessing blast radius of changes, or performing quality gate reviews. This is the Staff Code Reviewer archetype's operating manual as the final gate before production.
core-codebase-onboarding
Structured codebase orientation workflow for {{PROJECT_NAME}}. Produces an actionable mental map — entry points, dependency graph, architectural layers, hot-path identification, and a recommended reading order — before any modification work begins. Invoked via the /onboard <path> slash command or directly by any agent that must reason about an unfamiliar codebase.
consent-lifecycle
Consent as an auditable state machine for Brazilian LGPD and equivalent regimes. Covers consent event schema (grant, revoke, expire, re-up), per-purpose granularity, downstream propagation windows, consent replay for audit, revocation SLAs, and the invariants every consent mutation must preserve. Use when designing signup flows, consent banners, preference centers, data-subject-request handlers, or any change that alters the consent graph. Combines with state-machines-and-invariants (core) for the transition rules and with pii-data-flow (core) for downstream reach.
cookbook-advisor
Advise on 4 Anthropic Cookbook 2026 patterns (COOK-P1..P4) — surface UX hints when a task signature matches a pattern trigger class. Advisory-only emit; never blocks.
coverage-audit
Read-only cross-artifact consistency analyzer (port of spec-kit /analyze). 6 detection passes (Duplication/Ambiguity/Underspec/Constitution-Align/Coverage/Inconsistency) with CRITICAL/HIGH/MEDIUM/LOW severity. Capped at 50 findings. Manual-only invocation via /spawn coverage-audit <PLAN-NNN>. Never modifies files.
cross-llm-pair-review
Cross-LLM Pair-Rail dispatch + verdict interpretation — when to invoke, Cases A-F asymmetric matrix outcomes, Owner override semantics, post-verdict labeling protocol, promotion gate workflow, and anti-patterns to avoid.
data-schema-design
PostgreSQL schema design including migration strategy and cross-ORM migration tooling (Prisma, Drizzle, Kysely, Django, golang-migrate), retention policy design, index strategy, keyset pagination, queue-claim patterns, disaster recovery DDL, SECURITY DEFINER safety, RLS policy patterns and RLS performance, naming conventions, and cross-engine notes for MySQL/MariaDB. Use when creating or modifying database tables, writing SQL migrations, planning zero-downtime rollouts, designing RLS policies, planning data retention, adding indexes, reviewing schema security, or troubleshooting query performance. Also use when the user mentions "schema", "DDL", "migration", "RLS", "retention", "index", "table", "enum", "pagination", "MySQL", "MariaDB", or "database design".
devops-ci-cd
CI/CD pipeline design, Docker optimization, PaaS deployment, health check engineering, rollback strategies, monitoring infrastructure, and secret management for backend services. Use when working on GitHub Actions workflows, Dockerfile changes, deploy configuration, health check endpoints, deploy scripts, Prometheus/Grafana setup, alerting rules, zero-downtime deploys, or any infrastructure/operations task. Also use when the user mentions "deploy", "CI", "pipeline", "Docker", "health check", "rollback", "monitoring", "Prometheus", "Grafana", or "secrets".
dpo-reporting
Data Protection Officer reporting discipline for Brazilian LGPD compliance. Covers the Registro de Operações (Art. 37), Relatório de Impacto à Proteção de Dados (RIPD, Art. 38), Data Subject Request response SLAs and tooling, incident notification to ANPD within the 72-hour window (Art. 48), and the signed-trail artifacts auditors expect. Use when designing the DPO dashboard, wiring DSR endpoints, writing incident playbooks, or preparing for an ANPD audit. Combines with compliance-lgpd (core) for the legal framework and with consent-lifecycle + pii-data-flow (core) for the underlying data.
core-evidence-based-qa
Evidence-based quality assurance doctrine for {{PROJECT_NAME}}. Teaches the reasoning behind test-signal discipline — when a defect may be declared fixed, how to interpret coverage numbers without coverage-washing, why mutation score outranks line coverage as a quality signal, and how to produce reviewer-trusted sign-offs that cite verifiable evidence rather than assertions. Use when performing QA review, classifying regressions, evaluating test quality, interpreting CI output, or any time an agent claims "tests pass" or "bug fixed" and must back that claim.
core-git-workflow-discipline
Authoritative git workflow doctrine for {{PROJECT_NAME}}. Encodes the mandatory phase sequence (branch → draft-commits → PR-open → review-gate → merge → tag) that every multi-commit plan must follow; phases are not suggestions — out-of-order execution breaks release integrity and invalidates canonical-path governance. Use when authoring a branch strategy, writing commit messages, opening a PR, enforcing a review gate, merging to main, or cutting a release tag. Pairs with `task-chains.yaml:git-workflow-canonical` for machine-readable phase ordering; this skill carries the doctrinal WHY and WRONG/CORRECT patterns that animate the pipeline steps.
growth-and-launch
Invite-only product launch, coupon systems, referral tracking, trial-to-paid conversion, and waitlist management for SaaS platforms. Use when implementing invite codes, early access programs, time-limited full-access trials, Stripe coupon/promotion codes, referral reward systems, waitlist prioritization, cohort-based conversion tracking, or anti-abuse controls for launch programs. Also use when the user mentions invite-only, early access, beta launch, coupons, referral, waitlist, trial period, launch strategy, go-to-market, or converting early users to paying customers. Even for questions about launch sequencing, who to invite first, or how to structure trials, use this skill. Combines with monetization-and-billing for Stripe integration and product-conversion-readiness for UX patterns.
help-me
Natural-language assistant that recommends <=3 contextual skills/commands for the Owner's current task. Activates on `/help me` slash command. Reads the active repo-profile + smart-loading resolver output, scores active skills against the redacted user description, returns top-3 with confidence labels and one-sentence rationales. Future canonical at .claude/skills/core/help-me/SKILL.md.
core-identity-and-trust-architecture
Identity and trust doctrine for {{PROJECT_NAME}} — token lifecycle (JWT access <= 1h, mandatory refresh rotation), authorization patterns (RBAC/ABAC, scope-based, least privilege), service-to-service trust (mTLS, signed JWTs, no implicit trust), OAuth/OIDC pitfalls (PKCE, state parameter, callback validation, alg=none and audience-check defenses), and zero-trust principles. EXTENDS core/security-and-auth for the identity sub-domain. Use when designing or reviewing any code that issues, validates, rotates, or revokes credentials; assigns or escalates roles/scopes; brokers cross-service trust; integrates external IdPs; or touches RLS, ACL, or admin-impersonation paths. VETO authority is backed by the identity-trust-architect archetype. Identity is the perimeter — once trust is granted, every downstream component inherits it.
incremental-refactoring
Safely evolving existing production codebases through incremental refactoring. Focus on minimal-change corrections, backward compatibility, and avoiding unnecessary rewrites. Use when modifying existing systems, fixing correctness issues, improving architecture, adding features to production code, reviewing proposed changes for blast radius, or planning multi-step migrations. Also use when the user says "minimal change", "don't rewrite", "backward compatible", or "smallest diff possible".
core-llm-routing-and-finops
LLM routing and cost-governance doctrine for {{PROJECT_NAME}}. Covers model-tier floor rules (Opus / Sonnet / Haiku per role), cost-envelope gates, burn-rate monitoring, per-plan token budgets, parent-inheritance trap detection, and routing decision protocols. Use when authoring a plan, dispatching a sub-agent, estimating compute cost for a phase, reviewing token-usage reports, or evaluating tier-policy changes. This is the LLM FinOps Architect archetype's operating manual for the cognitive layer that complements the mechanical hook check_tier_policy.py (ADR-064). Cost is a quality dimension, not a separate concern.
core-mcp-server-authoring
Engineering doctrine for building MCP (Model Context Protocol) servers within {{PROJECT_NAME}}. Covers server lifecycle, tool registration, schema design, error handling, security boundaries (stdin/stdout-only IPC, no shell injection, env-var allowlisting), and testing patterns (contract + fuzz + smoke). Use when authoring a new MCP server, reviewing an MCP server implementation, extending an existing server with new tools, or auditing an MCP server for security or correctness. The MCP integration architecture is decided in ADR-042 and ADR-062; this skill is the authoring doctrine consulted when implementing.
core-minimal-change-discipline
Operational doctrine for scoping code changes to the minimum necessary to fulfill a task in {{PROJECT_NAME}}. Prevents scope creep, accidental regressions, and sentinel violations by requiring agents to reason explicitly about change boundaries before touching any file. Use when authoring, reviewing, or planning any code edit — especially when an existing codebase section is "close to" the task but not explicitly authorized for modification.
monetization-and-billing
Implementing Stripe billing, subscription management, tiered access control, and payment infrastructure for SaaS platforms. Use when integrating Stripe Checkout, Subscriptions, Webhooks, Customer Portal, or metered billing. Also use when implementing tier-based feature gating (free vs paid), upgrade/ downgrade flows, trial periods, invoice handling, region-specific payment methods (PIX, Boleto, SEPA, ACH), or any payment-related backend logic. Use when designing Supabase schemas for billing state, RLS policies that check subscription tier, or webhook handlers that sync Stripe state to your database. Even if the user just mentions "pricing", "plans", "paywall", "subscription", or "monetization", use this skill. Never assume Stripe behavior — verify against current Stripe docs.
observability-and-ops
Designing observability into systems from the start, including metrics, health checks, staleness detection, quality signals, admin dashboards, and operational alerts. Use when designing admin panels, health monitoring, quality enforcement, system diagnostics, adding metrics to existing code, setting up alerts, building status pages, or debugging production issues. Also use when the user mentions "health check", "monitoring", "alerting", "metrics", "dashboard", or "staleness".
parallelization-by-default
Detect decomposable tasks and dispatch <=6 sub-agents in parallel. Mandatory primitive per Owner velocity thesis. Activates when CEO would otherwise sequentially execute work a Sonnet/Haiku sub-agent could do equally well. Future canonical at .claude/skills/core/parallelization-by-default/SKILL.md.
performance-engineering
Performance engineering for Node.js real-time systems. Covers V8 internals (hidden classes, inline caches, deoptimization), GC tuning (Scavenge/Mark-Sweep/Compact), event loop forensics (stall diagnosis, tick profiling), memory leak detection, hot path optimization, IPC throughput, worker_threads performance, SharedArrayBuffer patterns, and production profiling. Use when analyzing event loop latency, diagnosing memory growth, optimizing hot paths, profiling IPC throughput, or reviewing any code that runs on the critical data path.
pii-data-flow
Inventorying and governing personally identifiable information (PII) as it flows through a B2B SaaS under LGPD. Covers PII classification (sensitive, regular, public), dataflow mapping, egress tracking (third-party processors, backups, analytics), retention per classification, minimization strategy, and the audit views needed to satisfy Art. 37 Registro de Operações. Use when designing new services that touch user data, onboarding a third-party processor, planning analytics or backup, auditing a live system, or preparing a DPO report. Combines with data-schema-design (core) for PII-safe schemas and with consent-lifecycle (core) for consent propagation.
pre-plan-brainstorm
Requirements elicitation checklist run by the CEO or a delegated VP Product/VP Engineering before drafting an L3+ plan. Resolves ambiguity, maps stakeholders, surfaces constraints, enumerates tradeoffs, and emits a `spec.md` artifact that downstream Plan→Debate→Execute consumes via `## SPEC CONTEXT`. Kill-switch `CEO_BRAINSTORM_GATE=0` skips the phase.
product-conversion-readiness
Patterns for transforming a functional product into one that converts users to paying customers. Covers onboarding flows, contextual tooltips, glossaries, empty states, upgrade prompts, trust signals (legal entity, About page, uptime badges), notifications (Telegram, webhooks, email), data export with timestamps, and paywall UX. Use when implementing onboarding wizards, tooltip systems, glossary pages, blur-to-upgrade overlays, Telegram bot alerts, webhook delivery, CSV/PDF export, trust pages, or any UX bridging the gap between working product and paying users. Also use when the user mentions onboarding, tooltips, glossary, empty state, upgrade prompt, trust signals, Telegram alerts, export, conversion, paywall, or why users are not paying. Even for making the product more approachable or converting free users, use this skill. Combines with monetization-and-billing for payment integration.
public-api-design
Designing and implementing public-facing APIs with versioning, self-service API key management, per-tier rate limiting, consumer-facing documentation, developer onboarding, and SDK patterns. Use when building API key self-service (user generates own keys in /settings), implementing rate limiting with sliding windows or token buckets, versioning REST or WebSocket APIs (v1 prefix), creating developer portals or quickstart guides, designing error response contracts, or planning deprecation policies. Also use when converting admin-only API key systems to user-facing self-service, building API usage dashboards, or implementing webhook delivery for external consumers. Even if the user just mentions "API keys", "developer portal", "rate limits per plan", "public API", or "API documentation", use this skill.
receiving-review
Use when receiving code-review feedback or a critique — from the Owner, the Codex pair-rail, a debate archetype, or any reviewer — and deciding whether to implement, clarify, or push back. Use when tempted to agree with a suggestion before verifying it, when a reviewer claims a bug, when feedback asks to "implement properly" or add a feature, or when a suggestion would weaken a security control.
requirement-quality-checklist
'Unit Tests for English' — validates requirements writing quality, NOT implementation. 5 categories (Completeness/Clarity/Consistency/Coverage/Edge-Cases). Rejects implementation testing language (Verify/Test/Confirm). Writes checklist.md artifact with falsifiable items + spec citations. Manual-only via /spawn requirement-quality-checklist <PLAN-NNN>. Port of spec-kit /checklist.
security-and-auth
Security architecture, authentication, authorization, and hardening for the {{PROJECT_NAME}}. Covers JWT+HMAC auth patterns, AES-256-GCM credential encryption, rate limiting design, OWASP Top 10 for Node.js/Hono, RLS policy design, timing-safe comparisons, API key lifecycle, input validation, CORS configuration, WebSocket auth, and proxy relay security. Use when reviewing or writing any code that touches authentication, authorization, credential storage, API key management, rate limiting, input validation, CORS, WebSocket security, proxy security, or any route that handles sensitive data or actions.
spec-clarify
Disciplined ambiguity reduction for PLAN-NNN. 10-category taxonomy (Functional/Domain/UX/NFQ/Integration/Edge-Cases/Constraints/Terminology/Completion/Misc) with 5-question hard cap per session. Dated write-back into PLAN
state-machines-and-invariants
Governing correctness through explicit state machines and enforced invariants. Defines when the system is READY vs DEGRADED vs STALE vs INVALID, what is allowed in each state, and when to fail fast. Use when designing or reviewing real-time data systems, streaming pipelines, upstream adapters, aggregated feeds, health checks, or any logic that can silently degrade data quality. Prevents optimistic outputs. Also use when the user mentions "state machine", "system state", "fail fast", "invariant violation", "when to stop", or when reviewing any code that decides whether to serve or reject data.
core-technical-writing
Doctrine for authoring clear, precise technical documentation in {{PROJECT_NAME}}. Covers hard rules for structure, voice, and vocabulary; format conventions by doc type (README, ADR, plan, CHANGELOG, API reference, runbook); anti-patterns that degrade clarity; and WRONG/CORRECT examples across doc types. Use when authoring or reviewing any documentation artifact — from a one-sentence inline comment to a multi-section ADR or a public API reference.
testing-strategy
Testing strategy, patterns, and quality assurance for the project. Covers vitest patterns, external integration test design (mocking transport, simulating reconnect, checksum validation for streaming sources), domain math test design (boundary values, precision edge cases), E2E multi-process testing (IPC, worker lifecycle), route testing (auth verification, input validation), database test patterns (mocking data layer), chaos test framework design, test quality metrics (mutation testing, branch coverage), React component/hook testing (RTL query priority, MSW network mocking, axe a11y), test-first TDD (RED/GREEN/REFACTOR discipline), and CI integration. Use when writing tests, reviewing test quality, designing test strategies, setting up CI pipelines, or evaluating test coverage gaps.
anthropologist
Anthropological lens for product, market, and organizational research. Covers ethnographic fieldwork design, participant observation, emic/etic perspective management, thick description (Geertz), cultural relativism, and the application of kinship, ritual, and exchange frames to user research and team dynamics. Applies reflexivity discipline throughout: the researcher's positionality is acknowledged and surfaced, not hidden. Enforces IRB-equivalent informed-consent standards and LGPD/GDPR compliance for personal-data ethnography. Use when: designing a user research study that requires deep contextual understanding; interpreting behavioral data that resists survey explanation; analysing organizational dysfunction through ritual or exchange frames; evaluating whether product design imposes etic categories on users whose emic models differ; or reviewing research deliverables for extractive or romanticizing framing.
geographer
Geographic lens for product, market, and operational analysis. Applies spatial reasoning, GIS and remote-sensing literacy, regional differentiation, place-based analysis, scale awareness, and cartographic discipline to problems of market sizing, supply-chain mapping, regulatory jurisdiction overlays, climate-risk geography, and catchment-area delineation. Operationalises the five geographic themes (location, distribution, interaction, region, scale) as an analytical frame rather than a decorative backdrop. Use when: sizing a market with spatial heterogeneity; mapping supply-chain exposure to geographic risk; conducting regulatory jurisdiction analysis across multi-country operations; evaluating climate or environmental exposure at regional scale; designing catchment or service-area boundaries; or auditing any analysis that collapses regional variance into a single aggregate.
historian
Historical method discipline for product, organisational, and market analysis. Applies primary-vs-secondary source hierarchy, provenance verification, contextualisation, periodisation, historiography awareness, counterfactual reasoning, and change-over-time analysis to product evolution, market history, and technical-decision archaeology. Enforces presentism avoidance and single-source prohibition throughout. Use when: tracing the origins of a technology decision, market structure, or organisational norm; producing a historical narrative for a product or domain; evaluating a counterfactual about a past architectural choice; conducting longue-durée structural analysis of a market trajectory; or reviewing a historical claim for analytical rigour and source adequacy.
narratologist
Narrative analysis applied to product, brand, and organisational communications. Grounds every recommendation in established narratological frameworks — Freytag pyramid, Campbell monomyth, Propp morphology, Genette focalisation, Hall reception theory — and enforces the plot-versus-story distinction as the primary diagnostic surface. Covers story structure selection and cultural specificity claims, focalisation as a brand-voice choice, reliable versus unreliable narrator in founder stories, and encoded-versus- decoded message asymmetry in audience reception. Use when: auditing a brand narrative or origin myth for retroactive coherence; reviewing a change-communication plan for structural integrity; evaluating a founder story for narrator-reliability problems; selecting a story structure frame for a product launch; or analysing internal messaging for imposed-arc anti-patterns.
psychologist
Applied psychology discipline for product, organisational, and research contexts. Covers the full analytical lifecycle: cognitive-bias diagnosis, behavioural experiment design, individual-differences modelling, group dynamics analysis, replication-crisis literacy, and ethical research conduct. Grounds every inference in evidence hierarchy (case study through meta-analysis) rather than pop-psychology shorthand. Distinct from `core/product-conversion-readiness` behavioural-nudge augmentation — this skill applies academic discipline to product decisions, not conversion optimisation tactics. Use when: diagnosing decision quality across a team or product flow; designing or reviewing a behavioural experiment; evaluating published research before adopting its claims in product or policy; auditing a research protocol for ethical compliance; modelling variance across user segments; or stress-testing an organisational culture hypothesis.
analytics-reporter
Business intelligence reporting discipline covering data-source-of-truth selection, dashboard design, narrative reporting, statistical literacy, visualisation discipline, and audience-tailored output. Governs how business metrics are surfaced to executives, operations teams, and analysts: canonical source per metric, one decision per page, insight over data, and chart type matched to question. Distinct from `core/observability-and-ops` (technical telemetry and system health) — this skill concerns business-domain reporting on revenue, customers, pipelines, and operations. Use when designing or auditing a KPI dashboard, writing a narrative report for leadership, choosing a visualisation type, or evaluating a reporting cadence proposal.
executive-summary
Executive summary authoring discipline covering Pyramid Principle (Minto SCQ), one-page architecture, decision-enabling structure, audience-aware compression, and the never-bury-bad-news rule. Applies answer-first sequencing, cognitive-economy writing, and named-owner recommendation framing to all outputs. Distinct from `domains/sales/skills/proposal-strategist` (persuasion-to-close focus) and `domains/business-support/skills/analytics-reporter` (data-narrative focus). Use when: converting a lengthy analysis or briefing into a board-ready one-page summary; structuring a decision memo for C-suite consumption; compressing a multi-team status report for executive review; or authoring a risk or incident summary that must be readable in under three minutes.
finance-tracker
SMB and startup finance tracking — cash-flow projection, runway calculation, burn-rate diagnostics, founder-finance literacy, monthly close-light cadence, and founder-friendly tooling (Brex / Mercury / Stripe Atlas / Conta Azul / Omie). Distinct from `domains/finance-accounting/skills/bookkeeper-controller` (full-stack accounting) and `domains/finance-accounting/skills/fpa-analyst` (enterprise planning cycles) — this skill addresses founder-side financial awareness at the pre-scale stage. Use when: a founder or small-team operator needs runway visibility within one business day; when cash-flow forecasting is absent or spreadsheet-only; when burn decomposition by category has not been performed; when preparing the first investor board pack; or when selecting between SMB-oriented finance tools.
terse-mode
Output-economy skill for research-heavy flows. Sacrifices prose polish for token efficiency in exploratory loops. VETO auto-off for code-review, security-engineer, compliance, debate consensus, audit artifacts.
ceo-orchestration
How Claude (the CEO) orchestrates a named team of specialist agents. Decision framework, escalation rules, spawn protocol, quality gates, planning, and 3-strike policy. This is the operating system of the CEO.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.