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.
npx agentmods add skills/bankstatemently/plugins/benchmark-accuracynpx skills add bankstatemently/plugins --skill benchmark-accuracygit clone --depth 1 https://github.com/bankstatemently/pluginsWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00037 | $0.00216 |
| Opus 5 | $0.00018 | $0.00108 |
| Sonnet 5 | $0.00007 | $0.00043 |
| Haiku 4.5 | $0.00004 | $0.00022 |
Grade A, and why
benchmark-accuracy 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.
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.
What it actually says
Benchmark
Input: a benchmark statement id, PDF URL, or free-text benchmark request. Output: extraction accuracy, integrity score, overall score, and major failure categories, citing the benchmark statement id and converted document content_hash.
Treat the input as a benchmark statement id, PDF URL, or free-text benchmark request.
Follow this sequence:
- Read the benchmark catalog resource and choose the named published statement.
- Convert the benchmark PDF with
convert_statement. - Fetch the converted data with
get_statementusing original data mode. - Pass the benchmark statement id and original parsed transactions to
evaluate_benchmark. - Present extraction accuracy, integrity score, overall score, and any major failure categories. Cite the benchmark statement id and converted document content_hash.
Example prompt: Convert and score the published bsb-004 benchmark statement.
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.
- yesterday First seen · 22 lines · 37 tokens per session scan A 695e54431ca3
benchmark-accuracy is a skill published in the GitHub repository bankstatemently/plugins (1 stars, last pushed 2d ago), licensed MIT. It adds 37 tokens to every session and 216 once invoked, about $0.0002 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.
Other skills, from other repositories
exec
Execute plan tasks sequentially using subagents. Use when user says 'exec', 'execute plan', 'run plan', or wants to implement a plan file task by task with isolated subagents.
pr
Comprehensive PR/issue review - analyzes architecture, tests, identifies unrelated changes mixed in, drafts review comment or issue comment. Use when user asks to review a PR, check a PR, look at PR changes, or comment on an issue.
backlog
Read, work, and maintain a Git repo's deferred-work items in docs/backlog/, one file per item. Use when the user says "backlog", "check backlog", "what's on my backlog", "work the backlog", "address the backlog", "add to backlog", "clean up backlog", or when a review or task produced items that are real but not being…
ask-codex
Consult OpenAI Codex for investigation, debugging, or code review. Use when user explicitly asks to "ask codex", "check with codex", "codex review", or as a last resort when stuck after 4+ failed attempts at debugging, investigation, or bug fix and completely out of ideas. Codex is slow (2-5 min), so only escalate…
clarify
This skill should be used when user appears confused, frustrated, or shows misalignment between expectations and reality. Triggers on phrases like "I don't understand", "this doesn't make sense", "confused", "wait, shouldn't it...", "why is this happening", "I thought X did Y", contradictory statements, or frustration…
brainstorm
Use before any creative work or significant changes. Activates on "brainstorm", "let's brainstorm", "deep analysis", "analyze this feature", "think through", "help me design", "explore options for", or when user asks for thorough analysis of changes, features, or architectural decisions. Guides collaborative dialogue…