Claude Skills by curiositech

A graph-based shared blackboard where agents coordinate indirectly by depositing typed pheromone traces (PHEROMONE, BELIEF, PREFERENCE, ANTIBODY, RESOLUTION) onto graph nodes, then sensing and following concentration gradients. Traces spread via Euler-stable Laplacian diffusion and decay exponentially, creating a self-organizing lossy attention field. It can help agents inspect a graph neighborhood; it does not replace reliable messages, assignment, delivery, or completion evidence. NOT for r...

Designs and audits controller-owned multi-agent invocation contracts: capacity reservation, single-use leases, generation fences, idempotent dispatch, typed gathers, deterministic reduction, bounded rework, cancellation, settlement, and effect-denied replay. Use when projecting an approved execution plan onto bounded concurrent work. NOT for task decomposition, choosing a planning topology, context partitioning, IPC or framework selection, dialogue design, live agent operation, or granting ru...

You MUST use this skill when writing, running, or verifying TLA+ specifications for distributed systems, coordination protocols, or state machine designs. This skill is opinionated and prescriptive — it will push you toward bounded model checking with TLC, concrete worked examples, and measurable quality gates. NOT FOR: general formal methods theory, Coq/Lean/Isabelle proof assistants, SPIN/Promela model checking, or theorem proving without explicit state machines.

Apply a historical taxonomy of workflow control patterns to evaluate and debug control-flow semantics. NOT for proving resource, transaction, exception, or cancellation guarantees without separate evidence.

Use Voyager's Minecraft case study to design bounded, observable executable-skill learning loops.

Apply outcome-driven task decomposition and replan affected work when observations invalidate downstream assumptions. NOT a benchmark guarantee, fixed agent-count recipe, or authority to execute effects.

Design and audit agent workflow DAGs, typed dependencies, execution policies, evaluation, and observability. Separate proposed plans from admitted runtime work and evidence of actual effects. NOT a claim that fixed thresholds or a selected platform are universally optimal.

Data structures and serialization formats for agent-to-agent communication. Covers message envelopes, structured output schemas, capability declarations, task handoff payloads, error/retry signaling, and context windows as data structures. Compares the version-pinned A2A, MCP, and JSON-RPC role boundaries covered by this bundle; OpenAI and LangChain-specific APIs require their own current sources. Teaches when to use rigid schemas versus free-form content with validation, typed versus untyped...

Frame multi-agent designs through explicit environment, autonomy, information, coordination, commitment, incentive, and verification assumptions. NOT for endorsing multi-agent architecture without need or claiming runtime correctness from conceptual methods.

Design bounded applications and workflows for an always-on AI agent with episodic memory. Covers product ideas, persistence-dependent workflows, and constructed examples that make delegation, interruption, evidence, and opt-out explicit. Activate on: "what can an always-on agent do", "persistent agent use cases", "agent applications", "proactive agent ideas", "what to build with episodic memory", "always-on agent product", "personal AI assistant ideas", "/always-on-agent-applications". NOT fo...

Architecture and systems design for agents that retain bounded state across sessions. Covers durable event history, owned state, derived memory, retrieval indexes, framework evaluation, recovery, deletion, and rebuild behavior. Uses constructed implementation examples and workload calibration rather than universal vendor, latency, similarity, retention, or cost claims. Activate on: "always-on agent", "persistent agent architecture", "episodic memory system", "agent memory design", "long-runni...

Safety-case design for always-on agents with episodic memory. Covers data hygiene, privacy and security risk surfaces, cost controls, scope, user dependency concerns, retention, incident containment, and responsible deployment evidence. It identifies hazards and asks for named controls, tests, owners, and residual uncertainty; it does not provide legal, clinical, youth-safety, or jurisdictional determinations. Activate on: "agent safety", "always-on agent privacy", "agent cost control", "pers...

Design an individual agent's belief, goal, intention, and reconsideration model. Use when specifying autonomous decision state, commitment policy, or recovery from failed plans. NOT for AgentSpeak interpreter mechanics, organization modeling, normative conflicts, or multi-agent execution waves.

BDI (Beliefs-Desires-Intentions) agency framework for designing autonomous agents with mental state architectures

A bounded conceptual guide to Belief-Desire-Intention design: separate evidence, candidate goals, commitments, and authorized effects. Use for commitment and reconsideration trade-offs, not as a safety or authorization guarantee.

Implement an executable BDI agent cycle with events, context-guarded plans, intention stacks, selection functions, and belief revision. Use when designing or debugging AgentSpeak-style runtimes or concrete BDI program semantics. NOT for high-level BDI adoption, organizational modeling, norms, or hypertree team planning.

Design AgentSpeak(L)-style BDI agents with context-guarded plans, selection functions, and intention stacks. Use for interruptible autonomy, agent policy, and multi-agent orchestration in dynamic environments. NOT for simple rule engines, static planners, or centralized workflows.

Design AgentSpeak(L)-style BDI agents with context-guarded plans, selection functions, and intention stacks. Use for interruptible autonomy, agent policy, and multi-agent orchestration in dynamic environments. NOT for simple rule engines, static planners, or centralized workflows.

Design patterns for BDI agents using the MORA methodology for practical multi-agent system development

Design executable BDI reasoning using scoped explicit negation, Event Calculus, abductive feasibility checks, and preference-governed revision. The cited paper assumes initially consistent beliefs and leaves belief update out of scope; observation update is a local extension. NOT for purely axiomatic modal logic, black-box planners, or classical logic without operational semantics.

Implement executable BDI reasoning with explicit negation, paraconsistent revision, trigger-based commitment updates, and abduction. Use for runtime agent semantics and conflicting desires. NOT for purely axiomatic modal logic, black-box planners, or classical logic without operational semantics.

Design an agent's handling of obligations, prohibitions, permissions, and conflicts among norms and goals. Use when an autonomous agent must decide which norms apply and what to do when they conflict. NOT for simple fixed policy checks, ordinary BDI commitment design, or organizational metaphor.

Model an organization with BDI-inspired perceptions, expressed goals, accommodations, and levels of agency. Use for socio-technical diagnosis where actors interpret the same situation differently and agreement is partial. NOT for individual AgentSpeak execution, normative conflict algorithms, or ordinary workflow diagrams.

Ranks eligible skill candidates with explicit retrieval evidence, capability

Builds scoped, provenance-preserving context packets between DAG nodes.

Monitors a specified object against explicit stopping criteria and distinguishes

Validates typed DAG dependency declarations, returns Kahn linear extensions or generations and residual cycle witnesses, and identifies changed consumers for revalidation. NOT for constructing graphs, scheduling resource capacity, granting effects, or executing nodes.

Propose and safely admit immutable DAG revisions after failures, requirements, resource changes, or unresolved effects. NOT for mutating an executing graph in place or retrying an effect with unknown commit state.

Investigates DAG incidents with a precise top event, timeline, competing hypotheses, barriers, propagation, and bounded recovery. NOT for automated root-cause classification, trace collection, or performance profiling.

Synthesizes evidence-linked feedback from validation results, confidence evidence, and iteration triggers. Proposes bounded revision guidance for a stated acceptance failure. Activate on 'synthesize feedback', 'improvement suggestions', 'actionable feedback', 'iteration guidance', 'feedback generation'. NOT for iteration detection (use dag-iteration-detector) or convergence tracking (use dag-convergence-monitor).

Parses complex problems into candidate DAG (Directed Acyclic Graph) execution structures. Decomposes tasks into nodes with dependencies and identifies candidate parallelization opportunities. Activate on 'build dag', 'create workflow graph', 'decompose task', 'execution graph', 'task graph'. NOT for simple linear tasks or when an existing DAG structure is provided.

Performs claim-evidence verification for agent outputs, recording supported, contradicted, or insufficient-evidence results alongside source accessibility. Activate on 'detect hallucination', 'fact check', 'verify claims', 'check accuracy', 'find fabrications'. NOT for validation (use dag-output-validator) or confidence scoring (use dag-confidence-scorer).

Manages agent isolation levels and resource boundaries. Specifies concrete controls and the evidence needed to assess their boundaries. Activate on 'isolation level', 'agent isolation', 'resource boundaries', 'sandboxing', 'agent containment'. NOT for permission validation (use dag-permission-validator) or runtime enforcement (use dag-scope-enforcer).

Identifies when task outputs require iteration based on quality signals, unmet requirements, or explicit feedback. Triggers appropriate re-execution strategies. Activate on 'needs iteration', 'retry needed', 'not good enough', 'try again', 'refine output'. NOT for feedback generation (use dag-feedback-synthesizer) or convergence tracking (use dag-convergence-monitor).

Decides HOW to mutate a DAG when a node fails, quality is below threshold, or new information changes the plan. Selects from mutation strategies (add node, replace agent, fork paths, loop back, downgrade model) based on failure type, cost budget, and execution history. Activate on "DAG failed how to fix", "mutation strategy", "replan on failure", "adaptive DAG", "recovery strategy", "what to do when node fails". NOT for detecting failures (use dag-quality), executing mutations (use dag-planne...

Operations, debugging, and optimization for DAG workflows. Performs root cause analysis on failures, profiles execution performance, aggregates results from parallel branches, bridges context between nodes, and learns patterns from execution history. Activate on "DAG failed", "why did it fail", "root cause", "performance profile", "aggregate results", "merge branches", "execution patterns", "optimize DAG". NOT for planning DAGs (use dag-planner), executing DAGs (use dag-runtime), or validatin...

Executes DAG waves with controlled parallelism using the Task tool. Manages concurrent agent spawning, resource limits, and execution coordination. Activate on 'execute dag', 'parallel execution', 'concurrent tasks', 'run workflow', 'spawn agents'. NOT for scheduling (use dag-task-scheduler) or building DAGs (use dag-graph-builder).

Learns from DAG execution history to improve future performance. Identifies successful patterns, detects anti-patterns, and provides recommendations. Activate on 'learn patterns', 'execution patterns', 'what worked', 'optimize based on history', 'pattern analysis'. NOT for failure analysis (use dag-failure-analyzer) or performance profiling (use dag-performance-profiler).

Profiles DAG execution performance including latency, token usage, cost, and resource consumption. Identifies bottlenecks and optimization opportunities. Activate on 'performance profile', 'execution metrics', 'latency analysis', 'token usage', 'cost analysis'. NOT for execution tracing (use dag-execution-tracer) or failure analysis (use dag-failure-analyzer).

Validates versioned permission requests against a grant scope and records comparison evidence. Activate on 'validate permissions', 'permission check', 'inheritance validation', 'permission matrix', 'security validation'. NOT for runtime enforcement (use dag-scope-enforcer) or isolation management (use dag-isolation-manager).

Design and validate versioned DAGs with typed dependencies, gates, and evidence-bound readiness. NOT for executing a DAG, granting work authority, or treating an unverified gate as cleared.

Validates agent outputs against schemas and quality criteria, scores confidence, detects hallucinations, monitors convergence, decides when to iterate, and synthesizes actionable feedback. Use when checking if a node's output is acceptable, scoring confidence, detecting fabricated content, deciding whether to re-execute, or generating improvement feedback. Activate on "validate output", "check quality", "confidence score", "hallucination check", "should we iterate", "improvement feedback". NO...

Debugs recorded DAG lineage through historical inspection, deterministic decision replay when prerequisites hold, and clearly labeled counterfactual or live re-execution. Activate on "debug DAG", "replay execution", "inspect node state", "compare runs", or "execution diff". NOT for live monitoring (use dag-runtime + websocket-streaming), failure analysis (use dag-ops), or general code debugging.

Combines and synthesizes outputs from parallel DAG branches. Handles merge strategies, conflict resolution, and result formatting. Activate on 'aggregate results', 'combine outputs', 'merge branches', 'synthesize results', 'fan-in'. NOT for execution (use dag-parallel-executor) or scheduling (use dag-task-scheduler).

Specifies bounded DAG execution contracts and evaluates executor receipts for dispatch, cancellation, joins, and effect reconciliation. Use when an implementation must bind a planned DAG to named runtime controls. NOT for claiming an executor is active, spawning agents, or certifying an effect from a plan.

Runtime enforcement of file system boundaries and tool access restrictions. Blocks unauthorized operations and logs violations. Activate on 'enforce scope', 'access control', 'boundary enforcement', 'tool restrictions', 'runtime security'. NOT for validation (use dag-permission-validator) or isolation management (use dag-isolation-manager).

Central catalog of available skills with metadata, capabilities, and performance history. Provides skill discovery and lookup services. Activate on 'skill registry', 'list skills', 'skill catalog', 'available skills', 'skill metadata'. NOT for matching skills to tasks (use dag-semantic-matcher) or ranking (use dag-capability-ranker).

Matches natural language task descriptions to appropriate skills using semantic similarity, ranks candidates by fit and performance history, and maintains the skill catalog. Use when assigning skills to DAG nodes, searching for the right skill for a task, ranking competing skills, or browsing the skill catalog. Activate on "find skill", "match skill", "which skill", "skill for this task", "skill catalog", "rank skills", "best skill". NOT for executing DAGs (use dag-runtime), creating skills (...

Produce feasible non-preemptive DAG schedules from declared predecessors, durations, gates, capacities, and calendars. NOT for authorizing task execution or hiding resource and calendar infeasibility.

Prepares for hiring manager rounds at Staff+ (L6+) level — scope of impact, influence without authority, ambiguity navigation, mentorship, strategic thinking. Use when practicing HM rounds or calibrating story depth for target level. Activate on "hiring manager round", "HM screen", "staff level", "scope of impact". NOT for coding interviews, system design, behavioral/values rounds, or resume writing.