coalesce-labs

Analyze user behavior patterns and cohorts using PostHog

Debug production errors using Sentry MCP tools. Searches issues, analyzes stack traces, identifies root causes, and suggests fixes. Use when the user mentions a Sentry error, production exception, stack trace, error monitoring, crash report, or unhandled exception.

Analyze distributed traces and performance issues with Sentry

Analyze the impact and scope of production errors

Analyze user segments, cohorts, and customer groups using PostHog data. Creates retention tables, compares segment behavior, and generates group breakdowns. Use when the user asks about customer segmentation, cohort retention, user group comparison, churn analysis, or funnel analysis by segment.

View key product metrics, KPIs, and conversion rates from PostHog

Reference template every phase-agent skill copies (CTL-448). The leading underscore on the directory name prevents the skill loader from picking this up — it is NOT a runnable skill, only a structural template the nine real phase agents (phase-triage, phase-research, phase-plan, phase-implement, phase-verify, phase-review, phase-pr, phase-monitor-merge, phase-monitor-deploy) clone and specialize.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Fix TypeScript errors with strict anti-reward-hacking rules. **ALWAYS use when** the user says 'fix type errors', 'fix typescript', 'type-check is failing', or when TypeScript compilation errors need to be resolved. Ensures runtime type safety — fixes root causes instead of silencing errors with casts.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Phase-agent wrapper that opens the pull request after implementation completes (CTL-449 Initiative 1 Phase 3). Delegates to `/catalyst-dev:create-pr` (which already auto-runs `describe-pr` and transitions Linear to `inReview`), then writes the PR number + URL into the phase signal file so the downstream `phase-monitor-merge` agent can read it without re-querying GitHub. Dispatched as a `claude --b

Phase-agent that fixes a failing verify verdict so the pipeline self-heals instead of stalling to needs-human (CTL-653). Reads `${ORCH_DIR}/workers/<ticket>/verify.json`, fixes the `findings[]` (every severity:"high" plus the regression_risk drivers) directly via Edit/Write, commits the remediation, and emits `phase.remediate.complete.<ticket>`. The scheduler's router then re-dispatches `verify` t

Set up leading vs lagging indicators for product decisions. Framework for metric selection and tracking.

Audit plugin health and find broken references in manifests, commands, agents, and skills

Analyze and reorganize a directory structure with safe reference updates

Fast browser automation CLI for AI agents. **ALWAYS use instead of Playwright MCP tools** for web testing, screenshots, form filling, and UI verification. Use when user says "open in browser", "check the site", "take a screenshot", "fill the form", "test the UI", or any browser interaction. Also use as a fallback when a task requires visual browser interaction that CLIs and APIs cannot handle (e.g

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Validate that implementation plans were correctly executed. **ALWAYS use when** the user says 'validate the plan', 'check if the plan was implemented correctly', 'verify the implementation', or after completing /implement-plan to confirm all phases were properly executed and success criteria met.

STEDII framework for selecting trustworthy experiment metrics. Ensures metric validity and reliability.

Create JTBD-based interview guides for user research. Structured questions for discovery interviews.

Research and catalog workflows from external Claude Code repositories

Connect MCPs for real-time tool integration

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Phase agent for the plan step of the 9-phase orchestrator pipeline (CTL-450). Wraps /catalyst-dev:create-plan and produces thoughts/shared/plans/<date>-<ticket>.md, then emits phase.plan.complete.<ticket>. Reads the prior research document from thoughts/shared/research/ as its prior-phase artifact. Spawned via plugins/dev/scripts/phase-agent-dispatch, which invokes it via slash command — hence `us

Systematically pull, categorize, and address all PR review comments — code change requests, questions, and suggestions. This skill fetches comments via gh api, groups them by file, implements fixes, handles disagreements diplomatically, and pushes a single commit. You should not try to handle PR review feedback manually — this skill ensures nothing gets missed. **ALWAYS consult this skill when** t

Create user journey maps and customer journey maps (dual mode)

Import and adapt a workflow from external repositories

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Phase-agent wrapper that drives TDD implementation from an approved plan (CTL-449 Initiative 1 Phase 3). Reads `thoughts/shared/plans/*-<ticket>.md`, delegates the red→green→refactor cycle to `/catalyst-dev:implement-plan`, commits each plan phase as it lands, and transitions the Linear ticket to `inProgress`. Dispatched as a `claude --bg` job by `phase-agent-dispatch`, which invokes it via slash

Phase agent for the verify step of the 9-phase orchestrator pipeline (CTL-450). NEW skill — has no canonical wrapper. Runs read-only adversarial verification against the implement-phase diff: tsc, tests, lint, security scan, reward-hacking scan, code review, test coverage, silent-failure hunt. Writes ${ORCH_DIR}/workers/<TICKET>/verify.json then emits phase.verify.complete.<ticket>. Reads phase-im

Deep competitive analysis + ongoing monitoring. Checks user research for competitor mentions, sales notes, existing analysis. Integrates with retention-analysis and user-research-synthesis.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Post-meeting effectiveness feedback and continuous improvement

Create new agents or commands using discovered patterns and templates

Create tickets via Linear MCP or generate formatted ticket text

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Phase agent for the research step of the 9-phase orchestrator pipeline (CTL-450). Wraps /catalyst-dev:research-codebase and produces thoughts/shared/research/<date>-<ticket>.md, then emits phase.research.complete.<ticket>. Reads triage.json from the worker dir as its prior-phase artifact. Spawned via plugins/dev/scripts/phase-agent-dispatch, which invokes it via slash command — hence `user-invocab

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.

Decide when to A/B test vs just ship. Framework for experiment planning and prioritization.

Analyze project milestone health with actionable insights, target date assessment, risk analysis, and specific recommendations

Create git commits autonomously for CI/automation (no user interaction). Non-interactive variant of /commit for use in CI pipelines, automated workflows, and background tasks. Never prompts the user.

Initial feature implementation from a PRD or feature description. **ALWAYS use when** the user wants to build a new feature, implement a PRD, create a first draft of code, or says things like 'build this feature', 'implement this PRD', 'code this up', 'create the initial implementation'. Also use when there's no existing codebase and user needs a standalone reference prototype.

Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.