ax-python-agent

Use when writing Python code with `axllm` for agents, child delegation, tools, MCP, citations, persistent playbook learning, stage instructions, runtime state, final typed responses, and direct-respond executor skipping.

Quality

86%

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

Quality

Content

87%

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

A lean, actionable overview with clean progressive disclosure and concrete code/API pointers. The main gap is workflow clarity: the skill advertises many capabilities but only demonstrates one core pattern, leaving the more complex operations unsequenced.

Suggestions

Add a short sequenced workflow (with a validation/check step) for at least one higher-stakes capability, e.g., harvesting run-end failures into the playbook or saving/restoring runtime state around long-running tasks.

Provide one minimal runnable example per major capability group (MCP client, tools, citations) or explicitly point to the matching entry under `examples/` so each 'When To Use' bullet has a concrete path.

Dimension	Reasoning	Score
Conciseness	The body is lean throughout — compact 'Package Facts' bullets, a minimal Core Pattern, a short API-surface list, and terse guardrails — with no explanation of concepts Claude already knows, matching the 'lean and efficient; every token earns its place' anchor.	3 / 3
Actionability	Provides an executable Core Pattern (`from axllm import agent; helper = agent(...); helper.forward(...)`), specific API symbols (`AxMCPClient`, `AxMCPStreamableHTTPTransport`, `RuntimeEnvelope`), and concrete guardrails ('Start from package examples', 'Use no-key examples for deterministic local checks'), meeting the 'fully executable code; specific examples' anchor.	3 / 3
Workflow Clarity	Structure exists (When To Use, Core Pattern, Guardrails) but only a single basic pattern is shown while the skill claims many capabilities (MCP, citations, playbook harvesting, runtime-state save/restore) with no sequenced steps or validation checkpoints for those, fitting 'steps listed but validation gaps; checkpoints missing or implicit' rather than the full-sequenced anchor.	2 / 3
Progressive Disclosure	The SKILL.md is a concise overview with well-organized sections and clearly signaled one-level-deep references ('Package API docs: `API.md` and `axir-api.json`', 'Runnable examples: `examples/`'), matching the 'clear overview with well-signaled one-level-deep references' anchor.	3 / 3
	Total	11 / 12 Passed

Description

85%

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

A specific, well-triggered description with an explicit 'Use when' clause and a clear niche. Its main weakness is heavy reliance on package-internal jargon that users are unlikely to say verbatim, capping trigger-term quality.

Suggestions

Add a few natural-language phrasings users would actually say (e.g., 'building LLM agents in Python', 'connecting MCP servers', 'adding tools to an agent') alongside the package-specific terms.

Soften or gloss jargon like 'direct-respond executor skipping' and 'stage instructions' with a plain-language equivalent so the trigger matches user intent rather than internal naming.

Dimension	Reasoning	Score
Specificity	Enumerates many concrete capabilities — "agents, child delegation, tools, MCP, citations, persistent playbook learning, stage instructions, runtime state, final typed responses, and direct-respond executor skipping" — rather than vague language, matching the 'lists multiple specific concrete actions' anchor.	3 / 3
Completeness	An explicit "Use when writing Python code with `axllm` for ..." clause provides the when, and the action plus enumerated capabilities provide the what, satisfying the 'clearly answers both what AND when with explicit triggers' anchor; it is not level 2 because the trigger is explicit rather than implied.	3 / 3
Trigger Term Quality	Includes some natural terms users would say ("writing Python code", "agents", "tools", "MCP", "citations"), but the bulk is package-specific jargon ("child delegation", "direct-respond executor skipping", "stage instructions") that users would rarely voice naturally, fitting the 'some relevant keywords but missing common variations' anchor rather than full natural-term coverage.	2 / 3
Distinctiveness Conflict Risk	Tied to the niche `axllm` package with a distinct feature set (MCP, playbook learning, citations), making it unlikely to fire for unrelated skills; this matches the 'clear niche with distinct triggers' anchor and is above the 'could still overlap' level 2.	3 / 3
	Total	11 / 12 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: ax-llm/ax
Path: packages/python/skills/ax-python-agent/SKILL.md
Commit: fe8baf8

Reviewed: 5 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.