ax-cpp-agent-context

Use when writing C++ code with `axllm` for deciding between context maps, trajectory context policy, offline optimization (ACE/GEPA), and memory recall for long-context agents.

Quality

75%

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 ./website/static/cpp/.well-known/agent-skills/ax-cpp-agent-context/SKILL.md

Quality

Content

68%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.

A concise, well-organized reference card for the axllm C++ package that respects token budget and stays factually dense. Its main weakness is that the routing decision it advertises — choosing among contextMap, contextPolicy, optimization, and recall — is never operationalized with criteria or a decision flow, and the lone code example is not executable standalone.

Suggestions

Add a short decision table or if/then list mapping task types to features (e.g., persistent corpus orientation → contextMap, within-run compaction → contextPolicy, prompt/program improvement offline → optimize, prior-run knowledge reuse → recall) so the skill's core routing task is actionable.

Make the Core Pattern snippet self-contained: show how llm is constructed (e.g., an axllm::LLM or no-key transport setup) or note the exact example file in examples/ to copy from.

Clarify where API.md, axir-api.json, axir-capabilities.json, and examples/ live (package root vs. skill bundle) so navigation to the deeper material is unambiguous.

Dimension	Reasoning	Score
Conciseness	The body is a lean bullet-structured reference: 'Language: C++', 'Package: axllm', 'Runtime Profiles: javascript-quickjs, python-pyodide' — no concept re-explanation or padding, and every fact is package-specific. It assumes Claude's competence throughout.	5 / 5
Actionability	The Core Pattern snippet is not self-contained ('helper.forward(llm, ...)' — llm is never constructed, no includes shown), and the skill's central task — 'Choose between contextMap, contextPolicy, optimization, and recall' — has no decision criteria mapping tasks to features. Real concrete guidance exists (specific symbols like 'axllm::AxGEPA', guardrails like 'Use no-key examples for deterministic local checks'), but key execution details are missing.	3 / 5
Workflow Clarity	Sections imply a rough flow (When To Use → Package Facts → Core Pattern → API Surface → Guardrails) and the guardrail 'Start from package examples for exact native syntax before inventing a new call shape' gives ordering advice, but there is no explicit sequence or checkpoint for the routing decision the skill exists to make. Not a destructive/batch skill, so no hard cap applies.	3 / 5
Progressive Disclosure	The skill is under 50 lines, well-organized into clear sections, and self-contained. It references 'API.md and axir-api.json', 'axir-capabilities.json', and 'examples/' that are not present in any skill bundle directory (references/, scripts/, assets/ do not exist), so whether these are package-relative or bundle-relative paths is ambiguous.	4 / 5
	Total	15 / 20 Passed

Description

71%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 distinctive, well-scoped description with an explicit 'Use when' trigger and four named capability areas. Its weaknesses are that the what is folded into the when-clause rather than stated separately, and trigger coverage is limited to canonical feature jargon without synonyms or natural user phrasings.

Suggestions

State the capability separately before the trigger, e.g., 'Routes long-context agent tasks in C++ (axllm) to the right feature: context maps, trajectory context policy, offline optimization (ACE/GEPA), or memory recall. Use when writing C++ with axllm and choosing how to supply context or optimize an agent.'

Add natural trigger synonyms users might type, such as 'context window management', 'prompt optimization', 'few-shot bootstrapping', or 'agent memory', alongside the canonical feature names.

Dimension	Reasoning	Score
Specificity	The description names the domain ('writing C++ code with axllm') and lists four concrete capability areas ('context maps, trajectory context policy, offline optimization (ACE/GEPA), and memory recall'). Not 5: a single verb ('deciding between') carries all four objects and coverage of what follows the decision is absent; not 3: several specific actions are named, more than the 1-2 of the midpoint anchor.	4 / 5
Completeness	An explicit trigger clause is present ('Use when writing C++ code with axllm for...') and the what — deciding among the four named features — is stated, but only embedded inside the when-clause rather than as its own statement of capability. Fits 'both what and when present, when/what could be more explicit' better than the level-3 anchor (what clear, when missing) or level-5 (both independently explicit).	4 / 5
Trigger Term Quality	Terms like 'context maps', 'trajectory context policy', 'ACE/GEPA', and 'memory recall' are what users of this package would say, but they are canonical feature names only — no synonyms, natural phrasing variations, or related terms a user might use instead. Matches the 'some relevant keywords but missing common variations' anchor.	3 / 5
Distinctiveness Conflict Risk	The scope is tightly pinned to C++ with the named package 'axllm' plus distinctive feature names like 'trajectory context policy' and 'ACE/GEPA' — a clear niche with triggers unlikely to fire for any other skill.	5 / 5
	Total	16 / 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: ax-llm/ax
Path: website/static/cpp/.well-known/agent-skills/ax-cpp-agent-context/SKILL.md
Commit: 3f630a0

Reviewed: 3 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.