update-success-rate

update-success-rate is a command for Claude Code from SawyerHood/dev-browser-eval. It costs 14 tokens per session (281 once invoked), scanned A, original, MIT.

A benchmark-reporting command for JSONL files, which are text files containing one JSON record per line. It checks the final result from each run and updates a comparison table.

In plain words
What is it for?
Use it to compare dev-browser, Playwright MCP, and Playwright skill runs by success rate, including both the percentage and the number of successful runs.
Why use it?
It removes the manual work of opening benchmark results, deciding which runs passed, and calculating percentages.

Command for Claude Code

Written for Claude Code: allowed-tools in frontmatter.

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/sawyerhood/dev-browser-eval/update-success-rate
Clone the repo
git clone --depth 1 https://github.com/SawyerHood/dev-browser-eval

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 update-success-rate

README.md
[![agentmods](https://agentmods.dev/badge/commands/sawyerhood/dev-browser-eval/update-success-rate.svg)](https://agentmods.dev/commands/sawyerhood/dev-browser-eval/update-success-rate)
Your own site
<a href="https://agentmods.dev/commands/sawyerhood/dev-browser-eval/update-success-rate"><img src="https://agentmods.dev/badge/commands/sawyerhood/dev-browser-eval/update-success-rate.svg" alt="Measured on agentmods" height="20"></a>
Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 281 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 $0.00014 $0.00281
Opus 5 $0.00007 $0.00140
Sonnet 5 $0.00003 $0.00056
Haiku 4.5 $0.00001 $0.00028

Measured 6d ago against content hash 7dbb8a8cee99, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

update-success-rate 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.

.claude/commands/update-success-rate.md · 34 lines

What it actually says

Update Benchmark Success Rates

Check the last line of all JSONL files in benchmark-results/ to verify task completion and update the success rate column in benchmark-comparison.md.

Steps

  1. Find all *.jsonl files in benchmark-results/
  2. Read the last line of each file (contains the final result JSON)
  3. For each run, check if:
    • The task completed successfully ("subtype":"success")
    • The average score is correct (should be 9.0)
  4. Calculate success rate for each method (dev-browser, playwright-mcp, playwright-skill)
  5. Update the table in benchmark-comparison.md with the Success Rate column

Success Criteria

A run is successful if:

  • The result shows "subtype":"success"
  • The result mentions an average score of 9.0 (not 8.0 or other values)

Output Format

Update the markdown table to include:

| Method | Time | Cost (USD) | Turns | Success Rate |

Where Success Rate shows the percentage and fraction, e.g., 100% (3/3) or 67% (2/3).

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 · 34 lines · 14 tokens per session scan A 7dbb8a8cee99

Subscribe to this mod's changes

update-success-rate is a command published in the GitHub repository SawyerHood/dev-browser-eval (23 stars, last pushed 8mo ago), licensed MIT. It adds 14 tokens to every session and 281 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.