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/bnet47/codexicon/initnpx skills add bnet47/codexicon --skill initgit clone --depth 1 https://github.com/bnet47/codexiconWhat 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 | $0.00024 | $0.01031 |
| Opus 5 | $0.00012 | $0.00515 |
| Sonnet 5 | $0.00005 | $0.00206 |
| Haiku 4.5 | $0.00002 | $0.00103 |
Grade A, and why
init 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.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Initialize the project
Announce: "I'm using init to turn the approved project context into a working development environment."
1. Establish context
Read the approved charter in agent_docs/briefs/ and inspect the repository for existing code, package manifests, and Git state. Do not overwrite an existing stack or working command without explaining the conflict.
If neither a charter nor equivalent explicit project context exists, stop and recommend $discover.
2. Resolve only missing choices
Ask only for decisions that cannot be inferred safely:
- language, framework, data store, and deploy target;
- whether an existing remote/repository must be preserved;
- whether the project needs external documentation or system context through MCP;
- any version or platform constraints.
When the user has no preference, recommend the smallest stack that fits the charter and explain the tradeoff briefly. Use $architecture-review if the choice is expensive to reverse.
3. Configure the repository
- Fill project identity in
AGENTS.md,README.md, and relevantagent_docs/files. - Confirm the project's actual licensing policy and replace the generic template copyright holder in
LICENSEwith the chosen owner when appropriate. - Replace the template reporting guidance in
SECURITY.mdwith the project's real private vulnerability route and accountable response owner before public or production use. - Complete the applicable boundaries and owners in
agent_docs/security.mdandagent_docs/operations.md; remove irrelevant template prompts rather than claiming they pass. - Create the actual stack skeleton and package metadata; remove template-only code that conflicts with it.
- Replace
scripts/setup.sh,dev.sh,lint.sh, andtest.shwith real commands. Replacedeploy.shonly when deployment is in scope. - Preserve
scripts/security.shandscripts/security.ps1as equivalent gates. Add ecosystem dependency, image, or infrastructure audits when applicable; do not remove the built-in credential scan without an equal or stronger local replacement. - Make an explicit static-analysis decision for the configured source languages and platform. Prefer repository-level GitHub CodeQL default setup when the repository is eligible and the stack is supported; otherwise select an appropriate SAST alternative or document why static analysis is not applicable. Record the scanner, language scope, triggers, alert owner, and query/rule policy in
agent_docs/security.md. Do not enable an external service or commit a reusable provider workflow without explicit authority and confirmed repository eligibility. - Support the developer platforms named in the charter. For native Windows projects, add equivalent
.ps1wrappers or document platform-neutral commands; each path must run the same underlying checks. - Preserve the lint/test verification receipt contract: call
.codex/hooks/codex_hook.py emit-success lint|testonly after the underlying check succeeds. If the project removes that contract, update the hooks, tests, anddocs/codex.mdtogether. - Replace the template validation workflow in
.github/workflows/ci.ymlwith CI that installs the selected stack and runs the same canonical lint, test, and security scripts. - Create
.env.examplewith names and comments only when environment variables are needed. Never create or read a real credential file. - Leave MCP absent unless the project needs external context. When needed, start from the commented project pattern, review the server and tool schemas, keep it disabled until the user trusts it, prefer read-only tools and scoped credentials, and record the trust boundary in
agent_docs/security.md. - Update
.gitignorefor generated outputs and local state. - Initialize Git only when no repository exists and the user chose a fresh local repository. Do not commit, add a remote, push, or open a PR unless separately requested.
- When Git is enabled, inspect any existing hooks configuration before running
scripts/install-git-hooks.shor.ps1. Preserve or deliberately integrate existing hooks; never silently overwrite a different hooks path.
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 · 68 lines · 0 tokens per session scan A b1eea05214ed
init is a skill published in the GitHub repository bnet47/codexicon (5 stars, last pushed 3d ago), licensed MIT. It adds 24 tokens to every session and 1,031 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 2026-08-31.
Other skills, from other repositories
make_plan
For external plan request scenarios, guides the Agent to request a clear, actionable, step-by-step plan from a stronger Agent via listagents and chatwithagent, emphasizing that the plan is executed by the requester, not by the consulted Agent.
当用户需要对PDF文件进行任何操作时,请使用此技能。包括从 PDF 中读取或提取文本/表格、合并多个 PDF、拆分 PDF、旋转页面、添加水印、创建新PDF、填写PDF表单、加密/解密 PDF、提取图片,以及对扫描版 PDF 进行 OCR 使其可搜索。如果用户提到 .pdf 文件或要求生成 PDF,请使用此技能。.
oma-scholar
Scholarly research companion using Knows sidecar spec (.knows.yaml). Generates, validates, reviews, queries, and compares structured research-paper sidecars, and fetches them from knows.academy. Use for academic literature search, survey synthesis, paper authoring assistance, and peer review with token-efficient…
oma-hwp
Convert HWP / HWPX / HWPML files to Markdown using kordoc. Extracts text, headings, tables, lists, images, footnotes, and hyperlinks. Use for Korean word processor files (Hangul), government documents, and AI-ready data preparation.
systematic-debugging
4-phase root cause debugging: understand bugs before fixing.
architecture-diagram
Dark-themed SVG architecture/cloud/infra diagrams as HTML.