CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

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

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./packages/python/skills/ax-python-agent/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

50%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 body is a dense, accurate-feeling API-behavior reference with clean section organization, but it reads as a spec dump: almost no executable examples beyond the two-line Core Pattern, heavy cross-language and TypeScript-relative detail, and no progressive offloading of the deep behavioral material. Claude would learn what the runtime does but not concretely how to build the skill's advertised tasks.

Suggestions

Add short executable code examples for the headline tasks (registering child agents, wiring an MCP client, streaming_forward with delta merging) instead of prose-only behavioral descriptions.

Move the Astra Session Work behavioral spec into a one-level-deep reference file and keep a 5-10 line summary inline, linking to it from a clearly signaled 'Advanced' section.

Strip TypeScript-relative comparisons and Java/Go/C++/Rust adapter details from the Python skill, and consolidate version-history notes ('since 25.0.0', 'before 25.0.0') into a dedicated 'Old patterns / deprecated' section.

DimensionReasoningScore

Conciseness

The body is information-dense with no padding of concepts Claude already knows, but it repeatedly anchors to TypeScript behavior ('as in TypeScript', ~6 times) in a Python skill, includes Java/Go/C++/Rust adapter detail irrelevant to Python authors, and carries inline version history ('the default since 25.0.0', 'as the ports did before 25.0.0') that the rubric penalizes outside a deprecated section. Mostly efficient but could be tightened.

3 / 5

Actionability

One executable snippet exists (the Core Pattern) plus scattered inline calls ('fn(name).namespace("crm")', 'add_child_agent(namespace, name, child)'), but most sections — especially the ~26-line Astra Session Work prose — describe semantics rather than provide executable how-to guidance for the skill's listed tasks (child agents, MCP clients, streaming).

3 / 5

Workflow Clarity

Topic-based sections with clear headers and a useful Guardrails section give coherent organization, but there is no sequenced workflow (e.g., start from examples → write code → verify with no-key transport) and no validation checkpoints; the skill is a multi-topic reference, so the simple-skill exception does not apply.

3 / 5

Progressive Disclosure

Section headers provide real structure and Package Facts signals where deeper material lives ('API.md', 'axir-api.json', 'examples/'), but no bundle files exist and the dense Astra Session Work behavioral spec is fully inlined in SKILL.md rather than offloaded to a one-level-deep reference file.

3 / 5

Total

12

/

20

Passed

Description

83%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.

A strong description with an explicit 'Use when' trigger anchored to a specific package name and a comprehensive enumeration of capability areas. Its weaknesses are jargon-heavy phrasing ('direct-respond executor skipping', 'stage instructions') and feature-noun lists rather than natural action verbs, which slightly blur trigger matching and specificity.

DimensionReasoningScore

Specificity

Enumerates many concrete capability areas ('agents, child delegation, tools, MCP, citations, persistent playbook learning... runtime state, final typed responses'), but these are noun phrases rather than actions, and terms like 'direct-respond executor skipping' are opaque — several specific items with minor gaps, not fully comprehensive action coverage.

4 / 5

Completeness

Explicitly answers both: 'Use when writing Python code with `axllm`' is a concrete when-clause, and the enumerated feature list states what the skill covers, with concrete trigger phrases (Python, axllm, agents, MCP, citations).

5 / 5

Trigger Term Quality

Includes natural terms users would say ('Python', 'axllm', 'agents', 'tools', 'MCP', 'citations', 'playbook'), but several phrases are internal jargon ('stage instructions', 'direct-respond executor skipping') and common synonyms are missing — good coverage, a few natural terms absent.

4 / 5

Distinctiveness Conflict Risk

The `axllm` package binding creates a clear niche with minimal wrong-skill triggering, but generic terms ('agents', 'tools', 'MCP') overlap with broader agent-framework skills — mostly distinct with minor overlap risk.

4 / 5

Total

17

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

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

Reviewed

Table of Contents

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.