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/gf-labs/ramp/calibrategit clone --depth 1 https://github.com/gf-labs/rampWrote 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.
[](https://agentmods.dev/commands/gf-labs/ramp/calibrate)<a href="https://agentmods.dev/commands/gf-labs/ramp/calibrate"><img src="https://agentmods.dev/badge/commands/gf-labs/ramp/calibrate.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00018 | $0.02423 |
| Opus 5 | $0.00009 | $0.01211 |
| Sonnet 5 | $0.00004 | $0.00485 |
| Haiku 4.5 | $0.00002 | $0.00242 |
Grade B, and why
calibrate scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
!`T=$(printf '%s' "$ARGUMENTS" | awk '{print $1}'); { [ -n "$T" ] && [ -f "$HOME/.claude/ramp/schemas/$T.md" ]; } || T="getting-started"; cat "$HOME/.claude/ramp/schemas/$T.md" 2>/dev/null || echo "NO_SCHEMA:$T"` How it starts
The opening of the file, as written. The whole thing — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context
Today's date: !date +%Y-%m-%d
Active topic (first word of the arguments if a schema exists for it, else getting-started):
!T=$(printf '%s' "$ARGUMENTS" | awk '{print $1}'); [ -n "$T" ] && [ -f "$HOME/.claude/ramp/schemas/$T.md" ] && echo "$T" || echo "getting-started"
Project (workspace home + evidence trail):
!P=$(cd "$(git rev-parse --show-toplevel 2>/dev/null || pwd)" && pwd -P); echo "$P (repo: $(basename "$P"))"
Topic schema (node map + detection-signal table):
!T=$(printf '%s' "$ARGUMENTS" | awk '{print $1}'); { [ -n "$T" ] && [ -f "$HOME/.claude/ramp/schemas/$T.md" ]; } || T="getting-started"; cat "$HOME/.claude/ramp/schemas/$T.md" 2>/dev/null || echo "NO_SCHEMA:$T"
Your saved graph (merge base — nothing here is ever downgraded):
!T=$(printf '%s' "$ARGUMENTS" | awk '{print $1}'); { [ -n "$T" ] && [ -f "$HOME/.claude/ramp/schemas/$T.md" ]; } || T="getting-started"; cat "$HOME/.claude/ramp/graphs/$T.md" 2>/dev/null || echo "NO_TREE_FILE:$T"
Scan signals (raw evidence the schema's detection table consumes):
!echo "CLAUDE.md lines: $([ -f CLAUDE.md ] && wc -l < CLAUDE.md | tr -d ' ' || echo 0)"; echo "git commits: $(git rev-list --count HEAD 2>/dev/null || echo 0)"; echo "claude sessions: $(find "$HOME/.claude/projects" -maxdepth 2 -name '*.jsonl' 2>/dev/null | wc -l | tr -d ' ')"; echo "project commands: $(ls .claude/commands/*.md 2>/dev/null | wc -l | tr -d ' ')"; echo "project hooks: $(python3 -c "import json;print(sorted(json.load(open('.claude/settings.json')).get('hooks',{}).keys()))" 2>/dev/null || echo none)"; echo "user mcp servers: $(python3 -c "import json,os;print(len(json.load(open(os.path.expanduser('~/.claude.json'))).get('mcpServers',{})))" 2>/dev/null || echo 0)"; echo "identity: $(git config user.name 2>/dev/null || echo unknown) <$(git config user.email 2>/dev/null || echo unknown)>"
Topic catalog (for the Explore next block):
![ -n "$CLAUDE_PLUGIN_ROOT" ] && python3 "$CLAUDE_PLUGIN_ROOT/ramp_core.py" catalog 2>/dev/null || echo "CATALOG_UNAVAILABLE"
Kernel CLI (the no-MCP write path):
![ -n "$CLAUDE_PLUGIN_ROOT" ] && [ -f "$CLAUDE_PLUGIN_ROOT/ramp_core.py" ] && echo "KERNEL_OK" || echo "KERNEL_UNAVAILABLE"
Existing placement worksheet (./.ramp/calibrate.md):
!cat .ramp/calibrate.md 2>/dev/null || echo "NO_CALIBRATE"
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.
- 5d ago First seen · 127 lines · 18 tokens per session scan B 73734d9f5633
calibrate is a command published in the GitHub repository gf-labs/ramp (2 stars, last pushed 1mo ago), licensed MIT. It adds 18 tokens to every session and 2,423 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 2026-08-31.
Other commands, from other repositories
design-tutorial
Interactive guided tour of Naksha — learn commands through real exercises, discover workflows, and get oriented in under 10 minutes.
coding-drill
Rapid-fire reps — pattern recognition, complexity-of-this-snippet, template recall — flashcard cadence; prioritizes your weak patterns.
coding-import
Import a LeetCode problem into your local library via WebFetch — no MCP, no login. Becomes a normal markdown problem you can practice or mock.
alt
Import an Exam Radar (OPTIMETA Alt plugin) export and fold its lecture-emphasis exam signal into the course index — radar.md, a lecture-emphasis column on coverage.md, and a gold-zone weakmap.
blind
Strategy-level blind drill on a known HW or example problem. User describes approach in prose (no math typing); Claude verifies against solution then saves clean reference to derivations/.
weakmap
Priority-ranked weakness report. No arg → fresh report from latest errors per pattern. With concept arg → patch latest report by adding the user-declared weakness, save as new timestamped file.