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 commands/tobiasblask/open-paper-machine/generate-plotgit clone --depth 1 https://github.com/TobiasBlask/open-paper-machineWhat 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 | $0.00024 | $0.00286 |
| Opus 5 | $0.00012 | $0.00143 |
| Sonnet 5 | $0.00005 | $0.00057 |
| Haiku 4.5 | $0.00002 | $0.00029 |
Grade B, and why
generate-plot scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
PB_SCRIPT="$(find ~/.claude/plugins -name paperbanana_direct.py -path '*/open-academic-paper-machine/*' 2>/dev/null | head -1)" && \ What it actually says
Generate Plot: $ARGUMENTS
Read the figure-engine skill first.
Steps
- Parse
$ARGUMENTS— extract the data file path and visualization intent - Read and validate the data file (CSV or JSON)
- PRIMARY: Use the direct Python API via
paperbanana_direct.pyscript:PB_SCRIPT="$(find ~/.claude/plugins -name paperbanana_direct.py -path '*/open-academic-paper-machine/*' 2>/dev/null | head -1)" && \ python3 "$PB_SCRIPT" plot \ --data '{"col1": [...], "col2": [...]}' \ --caption "Description of the plot" \ --output-dir figures/ \ --filename "fig_results_name.png" \ --iterations 3 - If the script is not found, FALLBACK: generate using Python (matplotlib/seaborn) with academic styling
- Save output PNG to
figures/in the workspace (300 DPI, tight layout) - Provide the LaTeX
\begin{figure}snippet ready for inclusion - Show the user the generated figure using the Read tool on the PNG path
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.
- 2d ago First seen · 26 lines · 24 tokens per session scan B aabda1b57ece
generate-plot is a command published in the GitHub repository TobiasBlask/open-paper-machine (18 stars, last pushed 4mo ago), licensed MIT. It adds 24 tokens to every session and 286 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
ars-lit-review
ARS academic-paper lit-review mode — annotated bibliography in paper format.
ars-revision-coach
ARS academic-paper revision-coach — peer-review roadmap or source-accounted real-committee response skeleton.
ars-3w
ARS deep-research three-way-scan mode — WHY / HOW / WHAT paper comparison.
ars-citation-check
ARS academic-paper citation-check mode — citation error report.
show_diff
Call the showdiff MCP tool to show me a side-by-side visual diff (it returns an image in the chat). If I passed a checkpoint sha in $ARGUMENTS, diff that checkpoint; otherwise show my current uncommitted changes.
ars-mark-read
ARS /ars-mark-read — record human-read signal for one or more citation keys.