文件操作
文件系统与批量文件任务
claude-code-plugins
melodic-software
"Explore the local codebase before making changes — read code, trace dependencies, scan git history, discover tests, and audit build and tool configuration, persisting an EXPLORE.md index plus sidecars. Dispatches a fresh-context subagent by default so the file reads stay out of the main conversation, with a documented inline escape hatch. Use when: 'explore the codebase', 'what exists for X' investigation, 'how does this work', 'where is X implemented', 'trace the dependencies', 'what tests cover this', or as step 1 before any code change."
claude-code-plugins
melodic-software
"Maintainer-facing drift-check and upstream sync for the firecrawl plugin's wrapper skill — tracks the firecrawl-cli npm release and the upstream SKILL.md source. Run only from a working-tree checkout. Actions: --check (default, read-only drift report) and the bare full update (gated npm upgrade + advisory skill-content integration). Not for consumers — consumers update via /plugin marketplace update."
claude-code-plugins
melodic-software
"Audit Claude Code permission GRANTS for portability and auto-mode durability — scans skill/command/agent frontmatter allowed-tools and settings.json/settings.local.json/user-global permissions.allow for interpreter-wildcard rules dropped in auto mode, hardcoded machine/user paths, and inert plugin self-grants. Use when: 'check permission rules', 'why was my allowed-tools grant ignored', 'audit allow rules', 'is this permission portable', after authoring a code-execution grant, or when a guarded helper is denied despite an allow rule. Report-only."
Clarify a dense artifact into something you can act on
melodic-software
"Faithfully clarify a dense, decision-heavy message so you can act on it — chunk it into one-decision-at-a-time, define the session's own jargon, and surface exactly what you must decide, with operative terms quoted verbatim and no loss of precision. Decision-dense content gets an HTML decision table with numbered rows. Use when: 'make this clear', 'clarify this', 'help me digest this', 'break this down', 'I can't parse this', 'what am I actually deciding here', 'this is a wall of text'. Empty argument targets the previous assistant response. This changes STRUCTURE, not altitude — a lossy plain-language / ELI5 drop is education:explain (if installed) instead. Sibling to adhd:shape, a standing session-wide posture; this is a one-shot reshape of one artifact."
claude-code-plugins
melodic-software
"Audit Claude Code configuration files — settings.json, settings.local.json, .mcp.json, hooks, plugins, permissions, environment variables — for correctness, security, and drift against current official docs. Use when: 'audit settings', 'check config', 'check for config drift', after a Claude Code update, or when permissions, hooks, plugins, or MCP servers may be misconfigured; pass --fix to apply auto-correctable findings with confirmation."
claude-code-plugins
melodic-software
"Read-only audit of a Claude Code INSTALLATION directory — the machine-scope ~/.claude tree plus ~/.claude.json — inventorying every file, separating what the product's own retention sweep already manages from what nothing manages, resolving what each number in a filename actually means before any process-liveness check, and detecting a deliberate or mid-experiment state before classifying anything as stale. Reports; never deletes. Use when: 'audit my .claude folder', 'what is in my ~/.claude', 'why is my Claude Code install so big', 'is anything stale in my Claude directory', 'does Claude Code clean up after itself', 'check cleanupPeriodDays', 'is this lock file dead', 'tidy my Claude Code install'. Not for: a repo's project-scope .claude config (use /claude-config:audit), or deleting anything (use /disk-hygiene:clean)."
claude-code-plugins
melodic-software
"One-shot plain-language explainer — drops any concept, code, error, architecture, or the previous assistant response to genuinely plain words (concrete analogy, zero jargon), then layers altitude up only on request (high-school, then peer level). Use when: 'I don't understand this', 'I don't get it', 'what does this actually do', 'what does this mean', 'explain simply', 'ELI5', 'rephrase that'. Empty argument targets the previous assistant response (anaphora), so 'I don't get it' needs no topic named. This changes ALTITUDE — trades precision for plain words; when the ask is instead to reorganize a dense message faithfully without losing precision (chunk it one-decision-at-a-time, define its jargon, surface the decisions), that is a STRUCTURE change, adhd:clarify (if installed), not an altitude drop. Sibling to education:teach — hand off there for multi-session coaching; this is a single-shot comprehension check, not ongoing tutoring."
iguana-strategy
Senpi-ai
IGUANA v1.0.0 — XYZ Macro Index Trend. Trend-follows the broad XYZ indices xyz:SP500 + xyz:XYZ100. No stock-picking, no commodities — just the broad indices, in whichever direction has the stronger 4-day move. The simplest possible XYZ exposure, closest thing to an index-fund equivalent (but 24/7 on Hyperliquid). Onboarding tier. Balanced DSL + 48h hard_timeout for weekend pricing-gap risk.
cuckoo-strategy
Senpi-ai
CUCKOO v1.0.0 — Copy-the-Copiers (meta-strategy follower). Auto-discovers the top strategies by realized performance (discovery_get_top_strategies), pulls each one's current positions, and trades the asset + direction the top performers agree on most — performance-weighted, so a stronger strategy gets more say (capped so one outlier can't dominate). Because the strategies it follows are themselves copy/algo strategies, Cuckoo copies the copiers. Distinct from Remora (operator-picked whales) and the individual trader- followers. NEW archetype #14. Let-winners-run DSL with a staleness cap.
marlin-strategy
Senpi-ai
MARLIN v1.0.0 — Order-Book Imbalance Momentum. A persistent resting-depth imbalance in the L2 book is a near-term pressure tell: bids ≫ asks = buy pressure, asks ≫ bids = sell pressure. Marlin enters in the imbalance direction WHEN 15m momentum and Smart Money agree — it times the entry on the book, then holds the move (it is NOT a scalper). Universe: BTC, ETH, SOL, HYPE. Wide "let winners run" DSL + 24h hard_timeout.
meerkat-strategy
Senpi-ai
MEERKAT v1.0.0 — Momentum-Event Sniper. Watches the Senpi momentum-event feed (leaderboard_get_momentum_events — the 4h rolling-window momentum / rank-jump events) and snipes the freshest, highest-tier events the moment they fire, entering in the momentum direction before the move is broadly known. Tier (event magnitude) + freshness gate the entry; SM alignment + rising volume are confirmation bonuses. Distinct from the Striker / rank-jump agents (Jaguar/Orca/Roach) which score a leaderboard universe. Let-winners-run DSL with a short hard_timeout.
AI Collaboration Standards
AsiaOstrich
Prevent AI hallucination and ensure evidence-based responses when analyzing code or making suggestions. Use when: analyzing code, making recommendations, providing options, or when user asks about confidence/certainty. Keywords: certainty, assumption, inference, evidence, source, 確定性, 推測, 假設, 來源, 證據.
AI Instruction File Standards Guide
AsiaOstrich
Create and maintain AI instruction files (CLAUDE.md, .cursorrules, etc.) with proper structure. Use when: creating AI instruction files, separating universal vs project-specific rules, configuring AI tools. Keywords: CLAUDE.md, cursorrules, windsurfrules, clinerules, AI instructions, system prompt, 指令檔案, AI 設定.
docgen
AsiaOstrich
"[UDS] Generate usage documentation from project sources"
derive
AsiaOstrich
"[UDS] Derive BDD scenarios, TDD skeletons, or ATDD tables from specifications"
Game Master — Core
Bobby-Gray
Deliver one in-character recap paragraph. Enter GM mode — no /gm prefix needed.
plan-converter
catlog22
Convert any planning/analysis/brainstorm output to .task/*.json multi-file format. Supports roadmap.jsonl, plan.json, plan-note.md, conclusions.json, synthesis.json.
analyze-with-file
catlog22
Interactive collaborative analysis with documented discussions, inline exploration, and evolving understanding. Serial execution with no agent delegation.
team-lifecycle-v2
catlog22
Unified team skill for full lifecycle - spec/impl/test. All roles invoke this skill with --role arg for role-specific execution. Triggers on "team lifecycle".
unified-execute-with-file
catlog22
Universal execution engine consuming .task/*.json directory format. Serial task execution with convergence verification, progress tracking via execution.md + execution-events.md.
collaborative-plan-with-file
catlog22
Serial collaborative planning with Plan Note - Multi-domain serial task generation, unified plan-note.md, conflict detection. No agent delegation.
ccw-help
catlog22
CCW command help system. Search, browse, recommend commands. Triggers "ccw-help", "ccw-issue".
team-tech-debt
catlog22
Unified team skill for tech debt identification and cleanup. All roles invoke this skill with --role arg for role-specific execution. Triggers on "team tech-debt", "tech debt cleanup", "技术债务".
rev-struct
P4nda0s
Reconstruct data structures by analyzing memory access patterns across functions