autoresearch:evals

autoresearch:evals is a command for Claude Code from uditgoenka/autoresearch. It costs 19 tokens per session (1,285 once invoked), scanned A, original, MIT.

A command that reads tab-separated results from repeated experiments and reports trends, flat periods, declining gains, and regressions. TSV is a plain-text table format with columns separated by tabs.

In plain words
What is it for?
Use it to analyze selected or discovered results files, detect metric trends and plateaus, and receive recommendations based on available columns.
Why use it?
It turns raw experiment records into an overview of what is improving, what has stopped improving, and where results got worse.

Command for Claude Code

Written for Claude Code: argument-hint in frontmatter. Also seen: names the AskUserQuestion tool.

About the project

autoresearch is an agent workflow that repeatedly changes a project, verifies a measurable result, keeps or discards the change, and continues iterating toward a goal. It is for autonomous improvement tasks in Claude Code, OpenCode, and OpenAI Codex across domains with mechanical success measures. The catalogue contains its commands, hooks, skills, plugin, agent, and instruction.

uditgoenka/autoresearch · 6,116 stars · on GitHub · udit.co

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 commands/uditgoenka/autoresearch/evals
Clone the repo
git clone --depth 1 https://github.com/uditgoenka/autoresearch

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for autoresearch:evals

README.md
[![agentmods](https://agentmods.dev/badge/commands/uditgoenka/autoresearch/evals.svg)](https://agentmods.dev/commands/uditgoenka/autoresearch/evals)
Your own site
<a href="https://agentmods.dev/commands/uditgoenka/autoresearch/evals"><img src="https://agentmods.dev/badge/commands/uditgoenka/autoresearch/evals.svg" alt="Measured on agentmods" height="20"></a>
Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,285 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 5.1 $0.00019 $0.01285
Opus 5 $0.00010 $0.00642
Sonnet 5 $0.00004 $0.00257
Haiku 4.5 $0.00002 $0.00128

Measured 6d ago against content hash 22582ed7a8bb, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

autoresearch:evals 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 6d 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

.claude/commands/autoresearch/evals.md · 119 lines

How it starts

The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.

EXECUTE IMMEDIATELY.

Parse Arguments

Extract from $ARGUMENTS:

  • Positional path to a specific TSV file
  • --format — output format: text (default console), json, md (markdown file)
  • --compare <path> — (v2.2.0 placeholder, not yet implemented)

Input Discovery

  1. If path provided → use that TSV directly
  2. If no path → scan current directory + autoresearch/*/ for *-results.tsv files
  3. If multiple found → AskUserQuestion: "Which results to analyze?" — list found files
  4. If none found → AskUserQuestion: "Provide path to results TSV"
  5. Also scan project root for v2.0.03 legacy TSV files (backward compat)

Parse TSV

  1. Read line 1: extract # metric_direction: higher_is_better|lower_is_better comment
    • If missing → infer from column names (metric/error_count → guess, or ask user)
  2. Read line 2: header row → detect available columns
  3. Read remaining lines: data rows
  4. Handle missing timestamp column gracefully (v2.0.03 compat)

Column Detection & Analysis

Activate analysis based on columns present in header:

Column Analysis
metric Trend direction, plateau detection (3+ flat iterations), diminishing returns, biggest single-iteration jumps
delta Per-iteration efficiency, cumulative improvement, effort-to-gain ratio
status Keep/discard rate, crash frequency, success streaks, failure clusters, longest winning streak
guard + guard-metric Guard failure rate, metric-improved-but-guard-failed analysis
severity Severity distribution (critical/high/medium/low/info), critical discovery rate per iteration
hypothesis + status Confirmation rate, investigation efficiency, most productive techniques
commit File hotspot analysis (cross-ref with git diff for kept commits), change size correlation
technique Technique effectiveness ranking
dimension Dimension coverage completeness (X/12)
candidate_label + judge_verdict Convergence speed, oscillation count
error_type Error category distribution, fix rate per category
classification New vs extension vs duplicate ratio, saturation curve
convergence_count Convergence trajectory

Read the full file on GitHub · 119 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. 6d ago First seen · 119 lines · 19 tokens per session scan A 22582ed7a8bb

Subscribe to this mod's changes

autoresearch:evals is a command published in the GitHub repository uditgoenka/autoresearch (6,116 stars, last pushed 23d ago), licensed MIT. It adds 19 tokens to every session and 1,285 once invoked, about $0.0001 per session on Opus 5. 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.