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/punt-labs/biff/talkgit clone --depth 1 https://github.com/punt-labs/biffWhat 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 | $0.00010 | $0.00296 |
| Opus 5 | $0.00005 | $0.00148 |
| Sonnet 5 | $0.00002 | $0.00059 |
| Haiku 4.5 | $0.00001 | $0.00030 |
Grade A, and why
talk 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 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.
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
Input
Arguments: $ARGUMENTS
Parse as: first token is the recipient username (strip leading @ if present), remaining tokens are the opening message (optional).
Examples:
kai→to="kai"kai hey, got a minute?→to="kai",message="hey, got a minute?"
Task
- Call
mcp__plugin_biff_tty__talkwith the parsed values. - Incoming messages from the partner appear on the status bar automatically (0-2s). No need to poll or call talk_listen.
- When the user wants to reply, send with
mcp__plugin_biff_tty__writeto the same user. - When the user says to stop, call
mcp__plugin_biff_tty__talk_end.
If $ARGUMENTS is "end", call mcp__plugin_biff_tty__talk_end directly.
Do not repeat or reformat tool output — it is already formatted by hooks.
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 · 29 lines · 10 tokens per session scan A 63050c5a0eac
talk is a command published in the GitHub repository punt-labs/biff (2 stars, last pushed 3d ago), licensed MIT. It adds 10 tokens to every session and 296 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 2026-08-31.
Other commands, from other repositories
meeting-listen
Play back a completed meeting summary as a voiced debate between personas.
meeting-hive
Run an autonomous PR/FAQ review meeting where four personas debate and reach consensus without user intervention.
vote
Assess whether a PR/FAQ should move forward with a structured go/no-go decision.
feedback
Incorporate feedback into PR/FAQ and redraft affected sections.
externalize
Generate an external press release from the PR/FAQ and CHANGELOG for a specific release.
import
Import an existing document and launch the full /prfaq workflow with extracted content.