Claude Agent Skill · by Affaan M

Agent Harness Construction

Install Agent Harness Construction skill for Claude Code from affaan-m/everything-claude-code.

Install
Terminal · npx
$npx skills add https://github.com/affaan-m/everything-claude-code --skill agent-harness-construction
Works with Paperclip

How Agent Harness Construction fits into a Paperclip company.

Agent Harness Construction drops into any Paperclip agent that handles this kind of work. Assign it to a specialist inside a pre-configured PaperclipOrg company and the skill becomes available on every heartbeat — no prompt engineering, no tool wiring.

S
SaaS FactoryPaired

Pre-configured AI company — 18 agents, 18 skills, one-time purchase.

$27$59
Explore pack
Source file
SKILL.md73 lines
Expand
---name: agent-harness-constructiondescription: Design and optimize AI agent action spaces, tool definitions, and observation formatting for higher completion rates.origin: ECC--- # Agent Harness Construction Use this skill when you are improving how an agent plans, calls tools, recovers from errors, and converges on completion. ## Core Model Agent output quality is constrained by:1. Action space quality2. Observation quality3. Recovery quality4. Context budget quality ## Action Space Design 1. Use stable, explicit tool names.2. Keep inputs schema-first and narrow.3. Return deterministic output shapes.4. Avoid catch-all tools unless isolation is impossible. ## Granularity Rules - Use micro-tools for high-risk operations (deploy, migration, permissions).- Use medium tools for common edit/read/search loops.- Use macro-tools only when round-trip overhead is the dominant cost. ## Observation Design Every tool response should include:- `status`: success|warning|error- `summary`: one-line result- `next_actions`: actionable follow-ups- `artifacts`: file paths / IDs ## Error Recovery Contract For every error path, include:- root cause hint- safe retry instruction- explicit stop condition ## Context Budgeting 1. Keep system prompt minimal and invariant.2. Move large guidance into skills loaded on demand.3. Prefer references to files over inlining long documents.4. Compact at phase boundaries, not arbitrary token thresholds. ## Architecture Pattern Guidance - ReAct: best for exploratory tasks with uncertain path.- Function-calling: best for structured deterministic flows.- Hybrid (recommended): ReAct planning + typed tool execution. ## Benchmarking Track:- completion rate- retries per task- pass@1 and pass@3- cost per successful task ## Anti-Patterns - Too many tools with overlapping semantics.- Opaque tool output with no recovery hints.- Error-only output without next steps.- Context overloading with irrelevant references.