Claude Skills by Snowflake-Labs

Deactivate Full Assist Mode — removes all CocoPlus feature flags. Data files (memory, meter history, grove patterns) are preserved. Idempotent.

Cortex-specific behavioral guidelines — always-on inner constraint layer for all CocoPlus personas

CocoBloom crystallization — converts a successful ship execution trace into a reusable skill file. Requires recent successful $ship. Invoked via $bloom crystallize.

Enter the Build phase of CocoBrew. Reads plan.md, invokes CocoHarvest to decompose the plan into parallel workstreams, assigns personas, generates flow.json stages, and begins execution via $flow run. Must have $plan completed and approved first.

Guided wizard to capture or update organizational Snowflake/Cortex standards into .cocoplus/context/<category>.md. Presents a menu of 6 standard types, collects answers via multi-turn dialogue, and commits the file.

Display the last N CocoCupper findings from .cocoplus/grove/cupper-findings.md. Usage: $cup history [n] where n defaults to 5.

Run a structured decision-capture dialogue before $plan — locks implementation choices (model, evaluation methodology, accuracy threshold, scope boundaries) into discuss.md to prevent silent decision drift during planning

Initialize a CocoFleet manifest for multi-process orchestration. Creates .cocoplus/fleet/[name]-manifest.json with a template containing meta fields, defaults, and one example instance. Usage: $fleet init [fleet-name].

Display the current CocoGrove ubiquitous language glossary in alphabetical order — shows all defined domain terms, definitions, aliases, and associated functions

Decompose an approved plan into parallel workstreams, assign specialist personas, classify stages as HITL or AFK (CocoLens), generate flow.json stages with checkpoints and dual-file state, and create per-stage prompt files. Includes adaptive parallelism, stall detection, shell identity injection, and consecutive failure escalation.

Context utilization monitor — background monitor that samples context window utilization via PostToolUse hook, surfaces advisory at 60% and critical warning with recovery decision matrix at 70%.

Spawn genuinely independent parallel subagents — each assigned a distinct perspective role (Skeptic, Optimist, Pragmatist, Domain Expert, Contrarian) — to analyze a topic from independent viewpoints and synthesize divergent findings into a decision-ready roundtable output

Analyze the impact of staged git changes against the committed Cortex function knowledge graph — shows which downstream functions are affected before you commit

CocoMeter accuracy learning — displays estimation history, current adjustment factor, sample size, and trend from the accuracy learning feedback loop. Usage: $meter accuracy

CocoOps — delivery intelligence dashboard. Dispatches $ops sub-commands for DORA metrics, sprint health, time-aware suggestions, and demo mode. Invoked via $ops [dora|sprint|suggest|demo|refresh].

CocoPlus configuration SSOT — $cocoplus sync propagates cocoplus.toml into downstream artifacts; $cocoplus migrate-config converts legacy safety-config.json. Invoked via $cocoplus sync and $cocoplus migrate-config.

Write a structured recovery snapshot to lifecycle/checkpoint.md before context reset — captures phase, stage, decisions, and pending items so $pod resume can restore precise context

CocoPull session archive search — full-text search across past sessions. Handles $pull search "<query>" with --since and --feature filters, and $pull index rebuild.

List all available CocoRecipe pipeline templates from the profile recipes/ folder and any project-local .cocoplus/recipes/ directory.

CocoReview — structured code review with six-severity findings vocabulary, progressive disclosure architecture, and universal anti-pattern baseline. Invoked via $review [file] [--complexity] [--security] [--architecture] [--language <lang>].

Relevance-ranked context loading — background subagent that fires before Build stages and persona invocations to inject ranked context from CocoGrove, CocoContext, Environment Inspector, Prompt Studio, and CocoDream.

Store a forward-looking idea with a trigger condition — the idea will surface automatically when the project reaches the specified state

CocoSentinel approval — records SHA-bound approval for a previously evaluated artifact. Invoked via $sentinel --approve [file]. Voids automatically if the artifact is modified after approval.

Activate CocoSpark brainstorm mode. Generates multiple divergent approaches, challenges assumptions, and explores what-if scenarios. Output is marked as exploration and does NOT modify spec or lifecycle artifacts. Use $spark to start, $spark-off to end.

Developer engagement observer — non-blocking, always-on observational layer that tracks Delegation Intensity, Review Depth, and Engagement Zone throughout a session. Summary surfaced at $ship and FULL checkpoints.

CocoWisdom insights — invokes Haiku synthesis to produce a structured analysis of project rejection patterns. Writes date-stamped insights file. Invoked via $wisdom insights.

Disable the Code Quality Advisor. Automatic SQL review stops. Existing findings are preserved.

List past CocoDream distillation sessions from .cocoplus/grove/dream-*.md files, showing timestamp, function count, and lesson candidate count for each session.

Disable Context Mode. Prompts are no longer prepended with CocoPlus context.

Generate documentation for Snowflake objects. Runs SQL to extract column metadata, creates a data dictionary, and builds a lineage diagram. Usage: $doc run [schema.table] or $doc run (for all accessible tables).

Scan the Snowflake environment: schemas, tables, views, stored procedures, Cortex endpoints, and user permissions. Writes timestamped snapshot to .cocoplus/snapshots/. Use $inspect --schema <name> for a schema-specific scan.

Pause the running CocoFlow pipeline. Sets a pause flag that stops new stages from starting. Running stages complete their current action before stopping.

Create an exploration git worktree on a new branch from the current phase. Allows isolated experimentation that can be merged back or discarded without affecting the main branch.

Manually capture a decision, pattern, or error resolution to the Memory Engine. Writes to .cocoplus/memory/ for cross-session persistence. Triggered manually or via post-tool-use hook.

List all available CocoPlus specialist personas with their triggers, models, modes, and locked tool sets. No preconditions required — works without $pod init.

Compare two prompt versions by running both against the same test inputs and showing output differences. Usage: $prompt compare [prompt-a-path] [prompt-b-path].

Roll back the CocoBrew lifecycle to a prior phase commit. Performs a git soft reset to the specified phase ID commit, discards subsequent commits, and updates AGENTS.md phase state. Use with caution — subsequent phase artifacts will be removed.

Set Safety Gate to normal mode. In normal mode, destructive SQL operations trigger a soft-gate confirmation prompt before execution. Recommended for development environments.

Review and acknowledge open Critical findings from a SecondEye report. Supports --hitl-only (acknowledge HITL-classified stages) and --blocking-only (acknowledge BLOCKING-classified findings only). Clears $build soft gate when all critical findings are acknowledged.

Exit CocoSpark brainstorm mode, optionally carry forward selected insights, and close the exploration session cleanly. Usage: $spark-off.

Routes Snowflake-related operations to Cortex Code CLI for specialized Snowflake expertise. Use when user asks about Snowflake databases, data warehouses, SQL queries on Snowflake, Cortex AI features, Snowpark, dynamic tables, data governance in Snowflake, Snowflake security, or mentions "Cortex" explicitly. Do NOT use for general programming, local file operations, non-Snowflake databases, web development, or infrastructure tasks unrelated to Snowflake.

Auto-routing skill loaded by the prompt filter hook. Routes Snowflake-related operations to Cortex Code CLI. Not for direct invocation — use $cortex-run instead.

ONLY load this skill when the user explicitly types $cortex-run or /cortex-run. NEVER load this skill from auto-routing hooks or keyword matching. For auto-routed prompts, use cortex-code:cortex-router instead.

Install Snowflake CLI and Cortex Code CLI. Use when cortex is not installed, when the user asks to set up Cortex Code, or when routing fails because the CLI is missing. Triggers: setup cortex, install cortex, cortex not found, CLI not installed, set up snowflake.