harness-agents

Add or use full agent harness runtimes like Claude Code, Codex, Pi, Cursor, Mastra, or ACP agents inside Agent-Native.

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/harness-agents/SKILL.md

Quality

Content

82%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content is highly actionable, lean, and well-structured for an internal integration skill, with executable TypeScript and precise identifiers throughout; its only gaps are the absence of explicit validation checkpoints and doc references that are named rather than linked.

Dimension	Reasoning	Score
Conciseness	The body is dense and technical, assuming familiarity with Agent-Native internals (no generic concept explanations), with every bullet carrying specific APIs, env vars, or routes; it sits at anchor 4 because a few sections (Adapter Guidance, ACP bullets) could be trimmed slightly, not quite the frictionless lean of anchor 5.	4 / 5
Actionability	It provides copy-paste-ready TypeScript (registerBuiltinAgentHarnesses, resolveAgentHarness, startAgentHarnessRun with a full options object, custom ACP resolve) plus exact env vars, SQL tables, and routes — fully executable guidance covering the common cases, matching anchor 5.	5 / 5
Workflow Clarity	The four numbered 'How' steps (register/resolve → start a turn → persist session → surface runs) form a clear sequence with guardrails via the 'Don't' list and sub-agent depth cap, but there are no explicit validation checkpoints, so it lands at anchor 4 rather than 5; the destructive/batch cap does not apply since this is a sandboxed integration workflow, not a destructive batch op.	4 / 5
Progressive Disclosure	Content is well-sectioned (Rule/Why/How/ACP/Adapter/Sandbox/Depth/Don't/Related) with one-level pointers to other docs ('See the Sandbox Adapters doc', 'See the Agent Teams doc') and a Related Skills list; no bundle files exist so structure is judged on organization, which is good but with inline doc references rather than linked files — anchor 4.	4 / 5
	Total	17 / 20 Passed

Description

65%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific and distinctive, naming concrete harness runtimes within a clear Agent-Native niche, but it omits any 'Use when...' trigger guidance, which caps completeness and limits its utility as a routing signal.

Suggestions

Append an explicit trigger clause, e.g. 'Use when adding or wiring Claude Code, Codex, Pi, Cursor, Mastra, or ACP agents into an Agent-Native app.'

Replace the generic 'Add or use' with more concrete verbs (register, resolve, start, persist) to lift specificity.

Add a synonym or file/CLI cue (e.g. 'agent harness', 'native coding agent', 'ACP') so natural-term coverage reaches anchor 5.

Dimension	Reasoning	Score
Specificity	It names the domain ('full agent harness runtimes') and six concrete runtimes (Claude Code, Codex, Pi, Cursor, Mastra, ACP), but the only verbs are the generic 'Add or use', so the actions themselves are minimal rather than comprehensive — matching anchor 3, not 4 which requires several specific actions.	3 / 5
Completeness	The 'what' is clear ('Add or use full agent harness runtimes ... inside Agent-Native') but there is no 'Use when...' or equivalent trigger clause, so per the guideline a missing explicit-when caps completeness at 3 (anchor 3: clear what, when missing).	3 / 5
Trigger Term Quality	It surfaces the product names a user would actually say ('Claude Code', 'Codex', 'Cursor', 'Mastra', 'ACP agents') plus 'Agent-Native', giving good keyword coverage; it sits above anchor 3 (which expects missing common terms) but below 5 because no synonyms or natural phrasings beyond the proper nouns are present.	4 / 5
Distinctiveness Conflict Risk	The niche is sharply defined — full agent harness runtimes integrated within Agent-Native, named explicitly — so it is unlikely to trigger for the wrong skill, matching anchor 5's clear niche with minimal conflict risk.	5 / 5
	Total	15 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: BuilderIO/agent-native
Path: .agents/skills/harness-agents/SKILL.md
Commit: a06e0ac

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.