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/callmetechie/fleet-manager/compose-downgit clone --depth 1 https://github.com/CallMeTechie/fleet-managerWhat 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 | $0.00033 | $0.00926 |
| Opus 5 | $0.00016 | $0.00463 |
| Sonnet 5 | $0.00007 | $0.00185 |
| Haiku 4.5 | $0.00003 | $0.00093 |
Grade A, and why
compose-down scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Compose Down
For a critical project this command MUST first confirm the server + project
with the user via AskUserQuestion, then run with the matching --confirm=<project>
token (which the user types). See the critical-gate handling below.
set -euo pipefail
source "${CLAUDE_PLUGIN_ROOT:-plugin}/commands/_compose-lib.sh"
REMOVE=0; PROJECT=""; SERVER_ARG=""; TOKEN=""
set -f # no globbing of unquoted $ARGUMENTS (review H2)
# shellcheck disable=SC2086 # intentional: split ARGUMENTS into positional params
set -- ${ARGUMENTS:-}
while [ $# -gt 0 ]; do
case "$1" in
--remove) REMOVE=1 ;;
--confirm=*) TOKEN="${1#*=}" ;;
--*) echo "Unknown flag: $1" >&2; exit 1 ;;
*) if [ -z "$SERVER_ARG" ]; then SERVER_ARG="$1"; else PROJECT="$1"; fi ;;
esac
shift
done
if [ -n "$SERVER_ARG" ] && [ -z "$PROJECT" ]; then PROJECT="$SERVER_ARG"; SERVER_ARG=""; fi
[ -n "$PROJECT" ] || { echo "Usage: /compose-down [server] <project> [--remove] [--confirm=<project>]" >&2; exit 1; }
[[ "$PROJECT" =~ ^[A-Za-z0-9_.-]+$ ]] || { echo "Invalid project name" >&2; exit 1; }
NAME="$(resolve_server "$SERVER_ARG")"
build_ssh "$NAME"; echo_target "$NAME"
is_scope_authorized "$NAME" docker_compose || {
echo "REFUSED: scope 'docker_compose' not authorized for '$NAME'." >&2; exit 1; }
resolve_docker_cmd "$NAME"; docker_precheck
ENTRY="$(compose_find_project "$PROJECT")"
[ -n "$ENTRY" ] || { echo "ERROR: project '$PROJECT' not found." >&2; exit 1; }
# Critical-gate: protected projects require the matching --confirm token (or env var).
if is_protected_resource "$NAME" compose_project "$PROJECT"; then
confirm_destructive "$PROJECT" "$TOKEN" || exit 1
fi
CFG_ARGS="$(compose_config_args "$(echo "$ENTRY" | jq -r '.ConfigFiles')")"
if [ "$REMOVE" -eq 1 ]; then ACTION="down"; else ACTION="stop"; fi
# shellcheck disable=SC2086
if "${FM_SSH[@]}" "$FM_DOCKER compose $CFG_ARGS $ACTION"; then RC=0; else RC=$?; fi
[ "$RC" -eq 0 ] || { echo "ERROR: compose $ACTION failed (exit $RC)." >&2; exit "$RC"; }
echo ""
# After `down`, the project is de-indexed and absence is success (review M1).
NEWSTATUS="$(compose_ls_json | jq -r --arg p "$PROJECT" '.[] | select(.Name==$p) | .Status' 2>/dev/null || true)"
if [ "$ACTION" = "down" ]; then
if [ -z "$NEWSTATUS" ]; then echo "Verdict: removed ('$PROJECT' no longer indexed)."; else echo "Verdict: down requested but still indexed as $NEWSTATUS — check above."; fi
else
echo "Verdict: stopped (status: ${NEWSTATUS:-unknown})."
fi
Critical-gate prose (mandatory): when is_protected_resource matches, before
running, use AskUserQuestion to confirm with the user: "Destructive action on
CRITICAL project '' on server ''. Proceed?" Only after an explicit
yes, re-run including --confirm=<project>. Never fabricate the token to skip the
question. (confirm_destructive refuses without the token/env var as a fail-safe
against accidental runs — see the design spec for the honest limits of this gate.)
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 · 69 lines · 33 tokens per session scan A 19d91948c988
compose-down is a command published in the GitHub repository CallMeTechie/fleet-manager (1 stars, last pushed 29d ago), licensed MIT. It adds 33 tokens to every session and 926 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 commands, from other repositories
domains
Apply the CLI runtime and safety contract before executing a command. Runtime --help is authoritative.
access
Apply the CLI runtime and safety contract before executing a command. Runtime --help is authoritative.
cleanup
Clean up TTS cache files to free up disk space.
deps-age
Analyze dependency freshness and maintenance activity.
release-prep
Automated release readiness validation.
review-pr
Comprehensive PR review checklist to ensure code quality, testing, and best practices.