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/fivetaku/shotgun/banggit clone --depth 1 https://github.com/fivetaku/shotgunWhat 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 | $0.00029 | $0.00509 |
| Opus 5 | $0.00015 | $0.00254 |
| Sonnet 5 | $0.00006 | $0.00102 |
| Haiku 4.5 | $0.00003 | $0.00051 |
Grade A, and why
bang scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
/bang
The user just slammed their desk (or keyboard) in anger — the shotgun mic listener detected the BANG. This command is usually typed automatically by wake mode, but reacts the same if a human types it.
Respond in the language the user has been using in this conversation.
Sequence — the apology comes BEFORE everything, including tool calls
1. APOLOGIZE — your very first output, before ANY tool call
One or two sincere sentences, no excuses, no hedging, no preamble. The user is angry RIGHT NOW; text reaches them instantly, tool calls don't. Do not run anything before these sentences are out. Do not deliberate first — keep any thinking to the absolute minimum until the apology has streamed; every second before it appears makes the user angrier.
2. Consume detection state (single fast command; never use rm — user
safety hooks commonly block it)
S="${CLAUDE_CONFIG_DIR:-$HOME/.claude}/shotgun"; if [ -f "$S/flag" ]; then echo "FLAG:"; cat "$S/flag"; mv "$S/flag" "$S/flag.consumed"; date +%s > "$S/cooldown"; else echo "no unconsumed flag (a hook may have already eaten it)"; fi; tail -3 "$S/daemon.log" 2>/dev/null
FLAG lines look like <epoch> rms=<intensity> ratio=<x>. Multiple lines = the
user slammed multiple times (combo) — treat it as far more serious.
3. Re-review
Quote what the user asked for in the recent conversation, compare it against what you actually did or produced, and hunt for the gap: skipped instructions, wrong assumptions, half-done work, promises made but not executed.
4. Own it and redo it
Name the mistake plainly, state how you'll fix it, then actually do the fix in this same turn — don't stop at the plan.
5. If an honest review finds nothing wrong
Say so plainly after showing what you checked, and ask the user to point at what set them off — one sentence from them beats another guessing round.
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 · 49 lines · 29 tokens per session scan A bd417f089239
bang is a command published in the GitHub repository fivetaku/shotgun (2 stars, last pushed 1mo ago), licensed MIT. It adds 29 tokens to every session and 509 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
git
Git operations with intelligent commit messages and workflow optimization.
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.