Agent Engineering

Build the agents themselves: prompt structure, tool definitions, evaluation loops, memory and multi-agent orchestration.

63 skills

1. 49

   DeepEval Skills

   SkillEvaluation, tracing and OpenTelemetry export for LLM apps, split into three non-overlapping skills.▲0Agent Engineering
2. 50

   Remember

   SkillCompresses past sessions through Haiku into daily summaries and reloads them at the next start.▲0Agent Engineering
3. 51

   Atomic Agents Skills

   SkillScaffolds Atomic Agents code where every agent, tool and context boundary is a validated schema.▲0Agent Engineering
4. 52

   Hugging Face Skills

   SkillModel choice, training, Spaces and SageMaker deployment as separate Hugging Face skills.▲0Agent Engineering
5. 53

   Discord Channel for Claude Code

   SkillRuns a Discord bot as a Claude Code channel, gated by a pairing code and a snowflake allowlist.▲0Agent Engineering
6. 54

   Telegram Channel for Claude Code

   SkillTurns a BotFather bot into a Claude Code channel with pairing, allowlists and per-group opt-in.▲0Agent Engineering
7. 55

   iMessage Channel for Claude Code

   SkillBridges Messages.app into Claude Code by reading chat.db directly, behind a handle allowlist.▲0Agent Engineering
8. 56

   LLM Application Dev

   SkillBundles the RAG, vector-index and prompting skills an LLM app needs, rebuilt around LangGraph.▲0Agent Engineering
9. 57

   LLM Finetuning

   SkillRuns fine-tuning as a gated lifecycle where the eval harness is built before any training config.▲0Agent Engineering
10. 58

    Agent Orchestration

    SkillTurns agent tuning into a measured loop with a baseline, classified failure modes and a rollback path.▲0Agent Engineering
11. 59

    Context Management

    SkillSaves and rehydrates project context between sessions under an explicit token budget.▲0Agent Engineering
12. 60

    Agent Teams

    SkillSpawns small parallel agent teams for review, debugging and feature work with strict file ownership.▲0Agent Engineering
13. 61

    PluginEval

    SkillScores a Claude Code skill on ten dimensions across static checks, an LLM judge and Monte Carlo runs.▲0Agent Engineering
14. 62

    Protect MCP

    SkillGates every tool call through a Cedar policy and signs the decision into a hash-chained receipt.▲0Agent Engineering
15. 63

    Skill Forge Essentials

    SkillThree behavioural skills aimed at the mistakes agents make rather than the ones humans make.▲0Agent Engineering