Home / miaodx / roboclaws · skills/eval-evolution/SKILL.md · GitHub

eval-evolution skillA

eval-evolution is agent-read markdown (skill) from miaodx/roboclaws: Prepare, run, review, and explicitly promote bounded Skill or existing-MCP Eval Evolution campaigns through the canonical eval facade..

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

What the file says

# Eval Evolution

Use this Skill for maintainer-owned optimization campaigns driven by frozen eval
evidence. `roboclaws.evals` remains the control plane. The optimizer and robot
trials are distinct OpenAI Agents SDK agents; a provider profile selects their
models and never changes the agent engine.

## Prepare

Freeze an `eval_evolution_campaign_v1` JSON manifest before the first optimizer
call. Bind one target kind and target, the baseline commit and target digest,
mutable paths, optimizer and robot provider/model identities, paired training
suites, the orchestrator-only sealed confirmation reference, quality and
minimum-improvement policy, runtime identity, and explicit turn/trial/token/
cost/time/retry ceilings.
The budget block also declares positive `optimizer_call_tokens`,
`optimizer_call_cost_usd`, `robot_attempt_tokens`, and
`robot_attempt_cost_usd` reservations. Each reservation is a frozen maximum for
one optimizer run or one robot attempt, not a post-hoc usage target.

Skill campaigns target exactly one `skills/<name>/SKILL.md`. Keep
`static-full` as the baseline. `no-skill` is a non-promotable negative control.
Do not mix Skill and MCP changes.

## Run
…

Read the whole file at its exact version.

How to install

Latest version
mdr add miaodx/roboclaws/eval-evolution@git:20260816.2788518
Exact content
mdr add miaodx/roboclaws/eval-evolution@sha256:a976866b6b4406f4

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_jyhbunigryx6zuoo.svg)](https://markdownregistry.com/a/art_jyhbunigryx6zuoo)

1 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260816.2788518 latest2026-08-16 2788518 3,688 BA view · diff
git:20260816.62caac32026-08-16 62caac3 3,070 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (3688 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

miaodx/roboclaws · 6 stars · license MIT · pushed 2026-09-24 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_jyhbunigryx6zuoo
GET https://markdownregistry.com/api/v1/resolve?ref=miaodx/roboclaws/eval-evolution
GET https://markdownregistry.com/api/v1/blob/a976866b6b4406f49da40044078303a18bf3111e5428a56d42608e59dcb28fec

Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

More from miaodx/roboclaws

AGENTS.md agents
miaodx/roboclaws · AGENTS.md
git:20260816.ecc907e · audit A · 6 stars
CLAUDE.md claude
miaodx/roboclaws · CLAUDE.md
git:20260626.e56711f · audit A · 6 stars
cloudml-eval-ops skill
miaodx/roboclaws · skills/cloudml-eval-ops/SKILL.md · Run frozen Roboclaws Eval Harness rows on CloudML with bounded parallelism, official cml lifecycle commands…
git:20260826.4cd1c7a · audit A · 6 stars
eval-harness skill
miaodx/roboclaws · skills/eval-harness/SKILL.md · Select and run Roboclaws validation, product, eval-suite, and live-agent eval rows from a plan, diff, or explicit…
git:20260918.a03741f · audit A · 6 stars
household-world skill
miaodx/roboclaws · skills/household-world/SKILL.md · Complete household-world goals through public household MCP tools.
git:20260918.c6faf47 · audit A · 6 stars
raw-fpv-visual-labeler skill
miaodx/roboclaws · skills/raw-fpv-visual-labeler/SKILL.md · Label visible cleanup-relevant movable objects from grouped RAW-FPV frames without creating executable cleanup handles.
git:20260622.3dd53c6 · audit A · 6 stars
report-performance-analysis skill
miaodx/roboclaws · skills/report-performance-analysis/SKILL.md · Use this skill whenever the user asks whether a Roboclaws live-agent report run is faster, wants to compare Agent…
git:20260728.cda130b · audit A · 6 stars
scene-gaussian-map-alignment skill
miaodx/roboclaws · skills/scene-gaussian-map-alignment/SKILL.md · Align scene Gaussian/splat, USD/mesh, and robot map assets into an honest digital-twin evidence workflow. Use when a…
git:20260816.3e0c2f2 · audit A · 6 stars
visual-result-showcase skill
miaodx/roboclaws · skills/visual-result-showcase/SKILL.md · Create blog, README, social, or demo-review visual showcases from completed Roboclaws run artifacts, especially…
git:20260816.39a87d9 · audit A · 6 stars

Every file in miaodx/roboclaws

Browse by kind, by grade A, or by owner.