Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add timwukp/agent-skills-best-practice --skill agentcore-harness-buildergit clone --depth 1 https://github.com/timwukp/agent-skills-best-practiceWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/timwukp/agent-skills-best-practice/agentcore-harness-builder)<a href="https://agentmods.dev/skills/timwukp/agent-skills-best-practice/agentcore-harness-builder"><img src="https://agentmods.dev/badge/skills/timwukp/agent-skills-best-practice/agentcore-harness-builder/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/timwukp/agent-skills-best-practice/agentcore-harness-builder"><img src="https://agentmods.dev/badge/skills/timwukp/agent-skills-best-practice/agentcore-harness-builder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00233 | $0.05699 |
| Opus 5 | $0.00117 | $0.02849 |
| Sonnet 5 | $0.00047 | $0.01140 |
| Haiku 4.5 | $0.00023 | $0.00570 |
Grade A, and why
agentcore-harness-builder scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 299 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AWS Bedrock AgentCore Harness Builder
Overview
A Harness is AWS Bedrock AgentCore's declarative, fully-managed way to run an agent. You hand AWS a JSON
configuration — model, system prompt, tools, memory, skills, limits — and AWS runs the agent loop (Strands under the
hood) inside a per-session Firecracker microVM with its own filesystem and shell. No container to build by default
(a custom image via environmentArtifact is an advanced option), no agent loop to write. You change behavior by
changing config, not redeploying code, and you can override model/prompt per invocation.
This skill builds a complete, best-practice Harness use case that exercises every AgentCore capability the user needs, wired correctly the first time. Harness is generally available (GA 2026-06-17, all AWS Commercial Regions where AgentCore is available, plus GovCloud US-West since 2026-08; only Payments remains preview), but it is still fast-moving and the real API shapes often differ from the published docs — this skill encodes the hard-won facts so you don't rediscover them through validation errors.
The two-plane mental model (internalize this first)
AgentCore has two distinct API surfaces. Confusing them is the #1 source of wasted time.
| Plane | What it is | How you call it |
|---|---|---|
| Control plane | Create/configure/inspect resources (harness, memory, runtimes) | boto3.client("bedrock-agentcore-control") — create_harness, update_harness, create_memory, … |
| Data plane | Invoke a running harness | boto3.client("bedrock-agentcore") — invoke_harness |
The agent-side SDK (pip install bedrock-agentcore) is a third thing: it's the library that runs inside a custom
Runtime container (BedrockAgentCoreApp, BrowserClient, MemorySessionManager). A Harness does not need it — the
managed harness loader image already wires the tools. You only touch the agent-side SDK if you drop down to Runtime
mode. See references/decision-guide.md.
What ships with it
39 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/harness.json.template 5.1 KB
- assets/iam_execution_role.json 5.0 KB
- assets/requirements.txt 507 B
- assets/skill.md.template 1.4 KB
- evals/evals.json 8.3 KB
- LICENSE.txt 11 KB
- LIVE_VALIDATION.md 13 KB
- references/advanced-config.md 6.2 KB
- references/browser-auth.md 6.9 KB
- references/code-interpreter.md 5.6 KB
- references/decision-guide.md 5.2 KB
- references/evaluations.md 7.3 KB
- references/gateway.md 37 KB
- references/gotchas.md 27 KB
- references/harness-config.md 9.2 KB
- references/identity.md 5.9 KB
- references/integrations.md 4.1 KB
- references/knowledge-bases.md 25 KB
- references/memory.md 9.8 KB
- references/model-and-prompt.md 8.3 KB
- references/observability.md 5.3 KB
- references/optimizations.md 10 KB
- references/payments.md 3.4 KB
- references/playground.md 2.1 KB
- references/policy.md 6.7 KB
- references/registry.md 17 KB
- references/runtime.md 16 KB
- references/skills.md 3.7 KB
- references/tools.md 8.6 KB
- references/versioning.md 3.3 KB
- scripts/create_harness.py 2.9 KB runs code
- scripts/invoke_harness.py 4.3 KB runs code
- scripts/preflight.py 6.0 KB runs code
- scripts/setup_observability.py 10 KB runs code
- scripts/test_offline.py 10 KB runs code
- scripts/update_harness.py 5.3 KB runs code
- scripts/validate_config.py 13 KB runs code
- scripts/wire_memory.py 7.5 KB runs code
- TEST_LOG.md 45 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 299 lines · 233 tokens per session scan A d98e3df0d49e
agentcore-harness-builder is a skill published in the GitHub repository timwukp/agent-skills-best-practice (10 stars, last pushed 3d ago), licensed MIT. It adds 233 tokens to every session and 5,699 once invoked, about $0.0012 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
bedrock
AWS Bedrock foundation models for generative AI. Use when invoking foundation models, building AI applications, creating embeddings, configuring model access, or implementing RAG patterns.
markitdown
Convert heterogeneous documents and selected URIs to Markdown with Microsoft MarkItDown for text analysis, search, and LLM/RAG ingestion. Covers safe local conversion, streams, Office/PDF/data formats, batch workflows, plugins, vision OCR, Azure extraction, and the official MCP server.
rag-architect
Designs and implements production-grade RAG systems by chunking documents, generating embeddings, configuring vector stores, building hybrid search pipelines, applying reranking, and evaluating retrieval quality. Use when building RAG systems, vector databases, or knowledge-grounded AI applications requiring semantic…
pgvector-semantic-search
Use this skill for setting up vector similarity search with pgvector for AI/ML embeddings, RAG applications, or semantic search. Trigger when user asks to: Store or search vector embeddings in PostgreSQL Set up semantic search, similarity search, or nearest neighbor search Create HNSW or IVFFlat indexes for vectors…
postgres-hybrid-text-search
Use this skill to implement hybrid search combining BM25 keyword search with semantic vector search using Reciprocal Rank Fusion (RRF). Trigger when user asks to: Combine keyword and semantic search Implement hybrid search or multi-modal retrieval Use BM25/pgtextsearch with pgvector together Implement RRF (Reciprocal…
rag-construction
Build RAG systems for construction knowledge bases. Create searchable AI-powered construction document systems.