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/wlfghdr/agentic-kbWrote 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/wlfghdr/agentic-kb/kb)<a href="https://agentmods.dev/commands/wlfghdr/agentic-kb/kb"><img src="https://agentmods.dev/badge/commands/wlfghdr/agentic-kb/kb/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/wlfghdr/agentic-kb/kb"><img src="https://agentmods.dev/badge/commands/wlfghdr/agentic-kb/kb.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 | $0.00020 | $0.01712 |
| Opus 5.5 | $0.00008 | $0.00685 |
| Sonnet 5.5 | $0.00004 | $0.00342 |
| Haiku 4.5 | $0.00002 | $0.00171 |
Grade A, and why
kb 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 6d 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 — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/kb — Knowledge Base
The user invokes /kb from any harness. Route to the kb-management skill.
Tool requirement. This prompt needs file + terminal tools (declared in the frontmatter
tools:list) to run file scaffolding, ritual scans, and git operations. On recent VS Code builds they are auto-selected from the frontmatter. If the chat session reports "no tools available" when this prompt runs, open the Chat view → gear/Configure Chat → Tools, enable the 13 built-in tools listed in the frontmatter above, and rerun/kb. This is a one-time per-session action.
Routing precedence
Evaluate in order and stop at the first match:
- No
.kb-config/layers.yamlanywhere in the workspace → hand off to thekb-setupskill (even if the user typed only/kbwith no args). Announce: "No KB detected — running setup." Then run the onboarding interview. - Explicit subcommand after
/kb(review,promote,publish,digest,todo,task,idea,develop,decide,note,brief,spec,release,incident,migrate,sync,diff,start-day,end-day,start-week,end-week,present,report,audit,status,setup) → route to that action perreferences/command-reference.md. - Product-management subcommand after
/kb—roadmap(orroadmaps) hands off to thekb-roadmapskill;journeys(orjourney) hands off to thekb-journeysskill. If the active layer has not declared the matchingroadmap:/journeys:config block in.kb-config/layers.yaml, refuse with a clear message that names the missing block, points the user at the skill'sreferences/config-schema.md, and offers/kb setupas the normal path for deciding which layer should own the artifact. - URL or pasted text → capture per the routing-mode rule (kb-management SKILL.md rule 12 +
references/capture-routing.md):- Default: no target named and no strong reflection signal → capture into the active layer (the anchor unless context already selected a different contributor-capable layer). Apply the five-question evaluation gate.
- Explicit: the user named a target layer in the invocation (e.g.
/kb team-observability <input>), OR acapture-routing:rule in.kb-config/layers.yamlmatches the input. Capture lands directly in the named layer; cite the matching rule (path + line or rule index) in the response. - Reflection-driven: no explicit target was named, but the input's content/source/context matches the strong-signal rubric in
references/capture-routing.mdfor a non-default contributor-capable layer. Propose the target, name the reason, and wait for human confirmation before mutating. Weak or ambiguous signals fall through to default. Do not soft-write to a staging area as a fallback; always offer the default as a one-word fallback.
- File path inside a known KB layer → layer-appropriate operation (review/update-topic/decide) on that file.
- Bare
/kb(no input) → run the triage scan below and present the result.
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.
- 6d ago Changed · -10 lines b3805c023a9e
- 1mo ago First seen · 99 lines · 20 tokens per session scan A ac515deb7c5a
kb is a command published in the GitHub repository wlfghdr/agentic-kb (10 stars, last pushed 6d ago), licensed Apache-2.0. It adds 20 tokens to every session and 1,712 once invoked, about $0.0001 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 2026-08-31.
Other commands, from other repositories
plan-sprint
Run a design sprint end to end — challenge framing, schedule, exercises, and prototype test plan.
dashboard-session-fork
Fork a session into a new one. Usage /dashboard:session-fork.
dashboard-session-hide
Hide a session from the dashboard list. Usage /dashboard:session-hide.
scout-meta-review
Run an interactive Scout meta-review — a system-level audit that sits above the individual session types. Checks whether sessions are running, mistake audit is trending well, proposals are flowing, KB files are healthy, and data-source coverage is consistent across session types. Runs in the current conversation.
scout-work
Interactive work session — walks through today's actionable items one at a time, presents a recommended action with draft content, and executes each one only with explicit approval. Runs in the current conversation (not as a background session).
hatch3r-roadmap
Sequence delivery phases over time into a dependency-ordered milestone plan with business and technical lenses, emitting a todo.md rollout schedule rather than design docs.