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.
git clone --depth 1 https://github.com/fockus/skill-memory-bankWrote 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/fockus/skill-memory-bank/config)<a href="https://agentmods.dev/commands/fockus/skill-memory-bank/config"><img src="https://agentmods.dev/badge/commands/fockus/skill-memory-bank/config/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/fockus/skill-memory-bank/config"><img src="https://agentmods.dev/badge/commands/fockus/skill-memory-bank/config.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00039 | $0.02139 |
| Opus 5.5 | $0.00016 | $0.00856 |
| Sonnet 5.5 | $0.00008 | $0.00428 |
| Haiku 4.5 | $0.00004 | $0.00214 |
Grade A, and why
config scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 14d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/mb config
Manage the project's execution pipeline.yaml — the declarative config consumed by /mb work. Defines roles → agents mapping, local workflow modes (workflow.default + workflows.*), per-item loops, severity gates, sprint context guard, review rubric, and SDD enforcement policy.
Skill bundle root
Every bundled helper below runs through the skill bundle root, never a bare scripts/… path (the
working directory is the user's project, where scripts/ is absent or belongs to someone else):
SKILL_DIR="${MB_SKILLS_ROOT:-${SKILL_DIR:-$HOME/.claude/skills/memory-bank}}"
[ -f "$SKILL_DIR/scripts/_lib.sh" ] || { echo "mb: skill bundle not found at $SKILL_DIR — set MB_SKILLS_ROOT" >&2; exit 2; }
Why pipeline.yaml?
/mb work <target> resolves a named workflow from pipeline.yaml, defaulting to medium (implement → verify → done, verifier once at plan end — review is off by default). Complexity presets: simple < medium < complex < governed (old names execution / governed-execution stay as aliases). Opt into review/judge per run (--review/--judge) or persist with review.enabled: true / <stage>.enabled: true; projects can also select full (the whole chain), full-cycle, planning-only, review-only, or custom loops with different max_cycles. Different teams need different defaults — review severity tolerance, max review cycles, role-to-agent mapping, protected-paths policy. Hard-coding these would lock the engine. pipeline.yaml makes them per-project and version-controlled.
Resolution
Effective config = first match wins:
<bank>/pipeline.yaml(project override, if present)references/pipeline.default.yaml(shipped default)
The shipped default is always present and self-validates. Projects do not need to run init until they want to override something.
Subcommands
| Subcommand | Description |
|---|---|
init [--force] |
Copy bundled default into <bank>/pipeline.yaml. Refuses if file exists unless --force is given. |
init --host <claude-code|codex|pi|opencode|cursor> [--preset simple|medium|complex|governed] [--cost premium|optimal|economy] [--model-premium X] [--model-mid Y] [--force] |
Write <bank>/pipeline.yaml from the default with workflow.default, cost and that host's model_profiles entry set (comments kept, result validated). |
show [--host H] [--cost C] [--preset P] [--verify V] |
Print the effective matrix: workflow steps + verify cadence, then per role agent → model → model_source → discipline, plus host notes and the project quality settings with their source (mb-profile.sh quality). |
show --raw |
Print the effective config file as-is (project override → default fallback). |
path |
Print absolute path to the effective config file. |
validate [yaml_file] |
Structural schema check (spec §9). Without an argument: validate the resolved file. With a file argument: validate that file directly. |
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.
- 14d ago Changed · +21 lines · +27 tokens per session b47562aae0b9
- 23d ago Changed · +10 lines da95fd471ac4
- 1mo ago First seen · 97 lines · 12 tokens per session scan A add57f45eaa7
config is a command published in the GitHub repository fockus/skill-memory-bank (29 stars, last pushed 3d ago), licensed MIT. It adds 39 tokens to every session and 2,139 once invoked, about $0.0002 per session on Opus 5.5. 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
research-verify
Verify existing research findings against independent primary sources. Upgrades confidence from 'sources agree' to 'independently verified.'.
story-continue
Resume an interrupted story from where you left off.
pause
Checkpoint the ticket in play so it can be resumed later.
unity-migrate
Plan and execute Unity version or render pipeline migration — identifies deprecated APIs, package compatibility, and executes migration steps.
unity-test
Writes missing tests and runs them via MCP. Identifies untested code, creates EditMode/PlayMode tests, executes via runtests, reports results.
spec
Interview and write a spec for a change to this codebase, then list the policies it touches.