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/fockus/skill-memory-bank/agreegit clone --depth 1 https://github.com/fockus/skill-memory-bankWhat 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 | $0.00013 | $0.01249 |
| Opus 5 | $0.00006 | $0.00624 |
| Sonnet 5 | $0.00003 | $0.00250 |
| Haiku 4.5 | $0.00001 | $0.00125 |
Grade A, and why
agree 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 3d 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.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agree: $ARGUMENTS
/mb agree <subcommand> [args] is the thin dispatch to scripts/mb-agree.sh — the ONLY writer of
<bank>/agreements.md (REQ-002). The model never edits the registry or the managed block by hand;
every mutation goes through this script so IDs stay monotonic, the format never drifts, and
parallel sessions serialize safely through the lock. Full protocol (what is/isn't an agreement, the
4 statuses, ADR routing, anti-examples): references/agreements.md.
0. Validate arguments
If $ARGUMENTS is empty, show the subcommand table below and ask which one to run. Do not guess a
subcommand from a bare statement.
1. CLI contract
| Subcommand | Args | Effect | Exit |
|---|---|---|---|
add "<statement>" |
[--supersedes N] [--adr NNN] [--source S] |
New active entry (default source user-confirmed, lazy-creates agreements.md from the template on first use); with --supersedes N atomically archives AGR-N as superseded and links it to the new entry |
0; 2 usage (empty/multiline statement); 1 invalid --supersedes target or damaged file |
defer N |
— | Move active AGR-N to ## Deferred |
0; 1 not active |
reject N |
— | Move active AGR-N to ## Archive marked rejected |
0; 1 not active |
question "<text>" |
— | Append - Q-NNN: <text> to ## Open Questions (no managed-block sync) |
0 |
resolve N |
— | Close open question Q-N |
0; 1 not found |
list |
[--all] |
Print ## Active (default) or the whole registry |
0 |
sync |
— | Rebuild the managed block in CLAUDE.md/AGENTS.md from the file (reconciles hand edits) |
0; 1 damaged block/file |
Statements are single-line — an embedded newline is a usage error (exit 2). Every mutating
subcommand takes the <bank>/.agreements.lock (same _lock_acquire/_lock_release idiom as
mb-work-progress-append.sh), writes via temp-file + mv, then re-syncs the managed block.
MB_AGREEMENTS=off (env or a .mb-config line) turns every subcommand into an explained no-op,
exit 0, zero writes.
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.
- 3d ago First seen · 88 lines · 13 tokens per session scan A d81005761fa2
agree is a command published in the GitHub repository fockus/skill-memory-bank (25 stars, last pushed 1mo ago), licensed MIT. It adds 13 tokens to every session and 1,249 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 2026-08-30.
Other commands, from other repositories
todo
The quality-gated task list: tasks with real descriptions, testable acceptance criteria, and evidence — a task only closes when the controller agrees it is done.
release
The pre-tag controller: version sync, changelog, clean tree, gate, and suite — every failure listed, the tag printed, never run.
security
The security pass: secrets (blocking), SAST, dependency vulns — plus the index's entry points to review from.
init
Install the formatters this repository needs, with every command visible before it runs.
vitaecontext-github
Audit or improve GitHub profile and repository SEO.
test
Run the test suite for the given package and summarize failures.