os-improvement-report

os-improvement-report is a skill for Claude Code from richfrem/agent-plugins-skills. It costs 81 tokens per session (919 once invoked), scanned A, original, MIT.

A reporting tool that turns logged agent improvement runs into a chart and text summary. It shows which changes were kept or discarded and how the best evaluation score changed over time.

In plain words
What is it for?
Use it to graph evaluation scores, review improvement cycles, and inspect results recorded by the experiment log, improvement ledger, or per-skill result files.
Why use it?
It makes progress across repeated agent tests easier to see than reading separate log files. It also helps reveal whether the agent is actually improving.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter.

Part of the agent-agentic-os plugin — 25 skills, 4 commands, 6 agents, 3 hooks shipped together

Good fit Use it to graph evaluation scores, review improvement cycles, and inspect results recorded by the experiment log, improvement ledger, or per-skill result files.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/richfrem/agent-plugins-skills/os-improvement-report
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.

Any agent
npx skills add richfrem/agent-plugins-skills --skill os-improvement-report
Clone the repo
git clone --depth 1 https://github.com/richfrem/agent-plugins-skills

Made for: Claude Code.

Or install agent-agentic-os, the plugin that ships this one along with the rest of its 25 skills, 4 commands, 6 agents, 3 hooks.

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 os-improvement-report

README.md
[![agentmods](https://agentmods.dev/badge/skills/richfrem/agent-plugins-skills/os-improvement-report/github.svg)](https://agentmods.dev/skills/richfrem/agent-plugins-skills/os-improvement-report)
Your own site
<a href="https://agentmods.dev/skills/richfrem/agent-plugins-skills/os-improvement-report"><img src="https://agentmods.dev/badge/skills/richfrem/agent-plugins-skills/os-improvement-report/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for os-improvement-report

Your own site · 80×15
<a href="https://agentmods.dev/skills/richfrem/agent-plugins-skills/os-improvement-report"><img src="https://agentmods.dev/badge/skills/richfrem/agent-plugins-skills/os-improvement-report.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 81 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 919 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00081 $0.00919
Opus 5 $0.00041 $0.00460
Sonnet 5 $0.00016 $0.00184
Haiku 4.5 $0.00008 $0.00092

Measured yesterday against content hash 43244c485baa, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

os-improvement-report 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 yesterday.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/generate_report.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

plugins/agent-agentic-os/skills/os-improvement-report/SKILL.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Loop Progress Report

Visual and text reporting on the agentic loop improvement cycle — across any plugin that maintains an improvement-ledger.md and results.tsv per skill.

The reference output is the autoresearch progress chart: green KEEP dots on a timeline, gray DISCARD dots, running-best step line, annotations showing what each improvement was. This skill produces the same chart for agentic-os and exploration-cycle-plugin improvement cycles.

Dependencies (Python 3.8+, pandas, matplotlib) are in references/detailed-reference.md.

What It Reads

Source Priority Content
context/experiment-log/index.md Primary All logged runs; filter result_type: numeric for KEEP/DISCARD/score data from orchestrator runs
context/memory/improvement-ledger.md Legacy fallback Eval score progression written by os-improvement-loop Stage 4.7; used if experiment log has no numeric entries
.agents/skills/*/evals/results.tsv Supplement Per-skill detailed eval score history

The experiment log is the unified source of truth for numeric results. The improvement ledger is a legacy format maintained for backward compatibility with older loop runs.

What It Produces

Output Description
context/memory/reports/progress_YYYYMMDD_HHMM.png Progress chart: KEEP/DISCARD timeline, running-best step line, change annotations
context/memory/reports/summary_YYYYMMDD_HHMM.md Text summary: baseline vs best, top hits by delta, survey effectiveness, north star trend

Execution Flow

  1. Read experiment log for numeric entries — run experiment_log.py summary, filter context/experiment-log/index.md for Result Type: numeric rows, and parse each linked file's KEEP/DISCARD verdict string. Fall through to Phase 1 if no numeric entries exist. Full parsing detail in references/detailed-reference.md.
  2. Check legacy data availability (fallback only) — if context/memory/improvement-ledger.md is missing or its Section 1 table is empty, tell the user no cycles have completed yet rather than running the report on an empty ledger.
  3. Run the report — invoke generate_report.py --project-dir ... --plugin-dir ... (optionally --skill <name>). Exits 0 and prints the chart path + text summary.
  4. Surface the output — report the chart path, print the text summary inline, and ask whether to open the chart image or show per-skill detail.
  5. Cross-plugin reporting (optional) — if tracking both agent-agentic-os and exploration-cycle-plugin, run the report once per plugin's project dir and concatenate the text summaries. Full commands in references/detailed-reference.md.

Read the full file on GitHub · 68 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. yesterday Changed · -130 lines 43244c485baa
  2. 9d ago First seen · 198 lines · 81 tokens per session scan A 95302c77d32c

Subscribe to this mod's changes

os-improvement-report is a skill published in the GitHub repository richfrem/agent-plugins-skills (6 stars, last pushed today), licensed MIT. It adds 81 tokens to every session and 919 once invoked, about $0.0004 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 2026-08-31.