skills-agents

Plugin skills-agents in poorvith-mp/skills-agents. Commands read from the repo's manifest — 1 month ago.

Skills (17)

• agent-architecture

  Design role decomposition, message passing, shared state and failure isolation across agents. Use when designing autonomous agent state machines, memory, or loops.

• agent-identity

  Design authentication, delegation, and trust verification across agents and identity sources. Use when establishing agent identity, auth, or provenance.

• agent-orchestration

  Sequence agents through an end-to-end pipeline, passing state between them, with pre-flight environment checks. Use when coordinating multi-agent systems, delegation, or subagent crews.

• agent-safety

  Defend against prompt injection, scope tool permissions, filter outputs, and fail closed when a check can't run. Use when implementing agent guardrails, sandboxing, or injection defense.

• automation-review

  Audit a proposed or existing automation for value, risk and maintenance cost, including the answer "don't build this.". Use when auditing automation ROI, failure points, or recommending 'don't build'.

• composio

  Wire agents to authenticated third-party SaaS tools through Composio connectors. Use when integrating AI agents with GitHub, Slack, Gmail, or Jira via Composio.

• create-skill

  Scaffold a new agent skill: frontmatter, trigger-rich description, structured body, references and a routing eval set. Use when authoring new agent skills, SKILL.md specs, or routing evals.

• document-generation

  Generate PDF, PPTX, DOCX and XLSX from code with real formatting, charts and tables. Use when generating PDF, Word, or Markdown documents from data and templates.

• mcp-builder

  Design, build and test MCP servers exposing tools, resources and prompts, including why a tool never gets called. Use when creating Model Context Protocol (MCP) servers, tools, or resources.

• n8n

  Build n8n workflows: node wiring, credentials, error branches and scheduling. Use when building visual workflow automations, webhooks, or nodes in n8n.

• prompt-engineering

  Write, test and optimise prompts into reliable production behaviour. Use when optimizing system prompts, few-shot examples, or reducing hallucinations.

• prompt-library

  Maintain reusable prompt libraries: versioning, tagging, benchmarking and deduplication across teams. Use when curating, organizing, or categorizing reusable prompt collections.

• rag-systems

  Build retrieval that works: chunking strategy, embeddings, hybrid search, reranking and retrieval evals. Use when building RAG pipelines, chunking, embeddings, or vector search.

• report-automation

  Merge multiple exports and unstructured sources into one consistent view, then schedule and distribute it. Use when automating scheduled metrics digests, executive summaries, or alerts.

• skill-linter

  Check a skill against house rules: description length, slug length, pointer scope and allowed frontmatter keys. Use when validating agent skills for length, triggers, evals, or spec rules.

• trigger-dev

  Build Trigger.dev background jobs: task definitions, retries, scheduling and observability. For visual workflows use n8n. Use when writing code-first background tasks or Trigger.dev.

• workflow-mapping

  Map a complete workflow tree: happy paths, branch conditions and failure states. Use when mapping processes into step-by-step logic, gates, or edge cases.

Similar plugins