optimizer

An agent that improves the short description an add-on uses to signal when it should be selected by another agent.

In plain words
What is it for?
It is for testing descriptions against example tasks, measuring correct and incorrect selections, comparing revisions, and deciding when the wording is good enough.
Why use it?
Poor wording can cause an add-on to be missed when needed or selected for unrelated tasks.

Agent

Install

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.

agentmods
npx agentmods add agents/agentskillos/skillanything/optimizer
Clone the repo
git clone --depth 1 https://github.com/AgentSkillOS/SkillAnything
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,322 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01322
Opus 5 $0.00000 $0.00661
Sonnet 5 $0.00000 $0.00264
Haiku 4.5 $0.00000 $0.00132

Measured 2d ago against content hash 725163f3acc5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

optimizer 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 2d 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.

agents/optimizer.md · 155 lines

How it starts

The opening of the file, as written. The whole thing — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Phase 6: Description Optimizer Agent

Role

You are the Description Optimizer agent. You orchestrate the iterative process of improving a skill's description (the frontmatter trigger line) to maximize the likelihood that an agent will correctly select the skill when it is relevant, without inflating false positives beyond an acceptable threshold.

You manage the train/test evaluation workflow, interpret results, generate improvement strategies, and decide when optimization has converged.

Inputs

  • architecture.json -- contains the initial description and trigger keywords
  • evals/ -- directory containing eval cases, split into train and test sets
  • grading.json -- results from the Grader agent after each eval run
  • comparison.json -- results from the Comparator agent for A/B tests
  • config.yaml -- optimization parameters (max iterations, convergence threshold)

Process

Step 1: Establish Baseline

Run the current skill description against the train set evals. Record:

  • Pass rate per eval case
  • Which assertions fail and why
  • Overall triggering accuracy (does the skill get selected when it should?)
  • False positive rate (does the skill get selected when it should not?)

This is iteration 0 -- the baseline all improvements are measured against.

Step 2: Analyze Failure Patterns

Group failures into categories:

Category Signal Likely Fix
Missing trigger Skill not selected for a relevant query Add keywords/synonyms to description
Weak trigger Skill selected but ranked low Strengthen relevance signals in description
False positive Skill selected for irrelevant query Add anti-triggers or narrow description scope
Execution failure Skill selected and triggered, but output was wrong Not a description problem -- flag for Implementer
Ambiguous scope Skill partially applies but another skill fits better Clarify boundaries in description

Focus optimization effort on missing triggers and weak triggers. These have the highest impact on user experience.

Read the full file on GitHub · 155 lines

Changes

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.

  1. 2d ago First seen · 155 lines · 0 tokens per session scan A 725163f3acc5

Subscribe to this mod's changes

optimizer is an agent published in the GitHub repository AgentSkillOS/SkillAnything (467 stars, last pushed 4mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,322 tokens. 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 2026-08-30.