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 skills/wlfghdr/agentic-kb/kb-setupnpx skills add wlfghdr/agentic-kb --skill kb-setupgit 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/skills/wlfghdr/agentic-kb/kb-setup)<a href="https://agentmods.dev/skills/wlfghdr/agentic-kb/kb-setup"><img src="https://agentmods.dev/badge/skills/wlfghdr/agentic-kb/kb-setup.svg" alt="Measured on agentmods" 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 | $0.00101 | $0.08695 |
| Opus 5 | $0.00051 | $0.04347 |
| Sonnet 5 | $0.00020 | $0.01739 |
| Haiku 4.5 | $0.00010 | $0.00869 |
Grade A, and why
kb-setup 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 — 394 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill: KB Setup
This skill is the single entry point for bootstrapping an agentic-kb workspace. It is idempotent: reruns add missing pieces, propose graph extensions, and never overwrite existing material without confirmation.
Scope boundary — install vs. init
This skill initializes the user's KB workspace. It does not distribute itself.
Two concerns, two tools:
| Concern | Who handles it | When |
|---|---|---|
Get kb-management + kb-setup + kb-operator into the harness |
Marketplace install or scripts/install.py |
Before this skill runs |
| Scaffold the user's layer repos, config, and workspace indexes | This skill | When the user types /kb setup |
When to invoke
- The user types
/kb setup. - The user says they want to set up, bootstrap, or onboard a KB workspace.
kb-managementdetects no.kb-config/layers.yamlin the current workspace and the user tries any/kbcommand.
Interactive question flow
The interview is goal-oriented, not feature-list-driven. Phase 1 collects the minimum admin facts the wizard cannot infer (workspace root and harness — frontloaded so the user knows where files will land before they invest in long answers); phase 2 asks the user about their world in their own language; phase 3 shows a derived plan and lets the user adjust before anything is written.
Ask each block in order. Stop and wait after each block for the user's answer before proceeding. Never ask the user to enumerate layers, features, contributor-mode flags, or scopes in phase 2 — those are derived in phase 3 and only adjusted there.
Phase 1 — Workspace and harness facts (short admin, frontloaded)
This phase runs first so the user always knows which directory the scaffold will land in and which harness will host /kb before they invest in open-ended answers. Wrong harness or wrong workspace root caught here is a one-line fix; caught after phase 2 it is sunk-cost frustration.
- Workspace root — absolute path; default current directory. Confirm with the user before continuing if the current directory looks accidental (e.g.
~/Downloads, a non-git directory the user did not expect). - IDE targets — multi-select from
claude-code,vscode,opencode,codex,gemini,kiro. Used byscripts/install.pyto decide where harness hooks are written (.claude/commands/,.github/prompts/,.agents/skills/, etc.). The right answer here prevents a silent "No KB detected" later. - Discovery pass — scan the workspace for existing KB repos, harness hooks, and likely layer candidates. Also probe for repo-as-OS structures (e.g.
work/signals/,work/missions/,org/<layer>/,CONFIG.yaml,CODEOWNERSwith policy directories). If a repo-as-OS structure is detected, surface it before phase 3 so the proposal can reuse existing repos and propose bridge defaults instead of inventing parallel structure.
What ships with it
40 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/adoption-stages.md 8.9 KB
- references/automation-levels.md 4.0 KB
- references/github-governance-profile.md 6.9 KB
- references/migration-guide.md 3.2 KB
- references/setup-flow.md 16 KB
- references/troubleshooting.md 2.9 KB
- templates/artifacts.yaml 3.7 KB
- templates/automation.yaml 1.8 KB
- templates/foundation-context.md 419 B
- templates/foundation-me.md 682 B
- templates/foundation-naming.md 520 B
- templates/foundation-sources.md 517 B
- templates/foundation-stakeholders.md 382 B
- templates/foundation-vmg.md 1.1 KB
- templates/github/agent-skills/kb-tracker-workflow/SKILL.md 9.1 KB
- templates/github/GITHUB_GOVERNANCE_SETUP.md 2.1 KB
- templates/github/ISSUE_TEMPLATE/bug.yml 1.0 KB
- templates/github/ISSUE_TEMPLATE/config.yml 45 B
- templates/github/ISSUE_TEMPLATE/content_update.yml 950 B
- templates/github/ISSUE_TEMPLATE/decision.yml 1.3 KB
- templates/github/ISSUE_TEMPLATE/feature.yml 923 B
- templates/github/ISSUE_TEMPLATE/feedback.yml 897 B
- templates/github/ISSUE_TEMPLATE/governance_change.yml 1.0 KB
- templates/github/ISSUE_TEMPLATE/idea.yml 850 B
- templates/github/ISSUE_TEMPLATE/roadmap_item.yml 1009 B
- templates/github/ISSUE_TEMPLATE/task.yml 802 B
- templates/github/labeler.yml 552 B
- templates/github/PULL_REQUEST_TEMPLATE.md 1.3 KB
- templates/github/workflows/kb-github-governance.yml 3.1 KB
- templates/kb.instructions.md 1.3 KB
- templates/kb.prompt.md 6.8 KB
- templates/layers.yaml 4.0 KB
- templates/org-kb-AGENTS.md 631 B
- templates/org-kb-README.md 718 B
- templates/personal-kb-AGENTS.md 1.1 KB
- templates/personal-kb-README.md 1.5 KB
- templates/presentation-template.html 30 KB
- templates/team-kb-AGENTS.md 881 B
- templates/team-kb-README.md 1.5 KB
- templates/workspace-AGENTS.md 758 B
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 · 394 lines · 101 tokens per session scan A fcde67e8ba25
kb-setup is a skill published in the GitHub repository wlfghdr/agentic-kb (10 stars, last pushed 19d ago), licensed Apache-2.0. It adds 101 tokens to every session and 8,695 once invoked, about $0.0005 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 skills, from other repositories
frame-macos-notification
拟真 macOS 通知 banner + app icon + 标题正文, 适合 video overlay / 产品发布预告.
deck-course-module
暖纸背景 + Playfair, 左侧学习目标常驻, 含 MCQ 自测页.
deck-pitch
10 页融资 deck, 白底 + 蓝紫渐变 hero, traction 柱状, $X.XM ask.
argent-tv-interact
Control and inspect TV apps via argent — Apple TV (tvOS), Android TV (leanback), and Amazon Fire TV (Vega). Boot the target, read focus, navigate with the D-pad remote, type, screenshot, and on Vega debug the JS runtime (evaluate, console logs, network inspector). Use when a task targets a TV (runtimeKind "tv", or…
review-offered-task
Review a task that has been offered to you and decide whether to accept or reject it.
aidd-dev:08:for-sure
Iterative agent loop that tracks attempts and retries until a success condition is met. Use when the user says "for sure", "make sure", "keep trying until", "loop until done", "don't stop until", or needs guaranteed completion of a task with explicit success criteria.