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/avery320/codex-vault/comet-knowledgenpx skills add Avery320/codex-vault --skill comet-knowledgegit clone --depth 1 https://github.com/Avery320/codex-vaultWhat 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 | $0.00071 | $0.00512 |
| Opus 5 | $0.00036 | $0.00256 |
| Sonnet 5 | $0.00014 | $0.00102 |
| Haiku 4.5 | $0.00007 | $0.00051 |
Grade A, and why
comet-knowledge 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.
What it actually says
COMET Knowledge
Use COMET's MCP tools as the source of truth for remembered vaults and the active workspace.
- When the user asks to open the knowledge workspace, or directly invokes COMET or this skill without another request, immediately call
show_comet_explorer. Pass the current Codex project root asproject_pathwhen it is known. Do not stop after acknowledging the invocation. - Start with
list_vaultswhen the active vault is unclear. Callget_workspace_infoafter a vault is active to confirm its path, counts, and access mode. - Use
select_vaultto switch an existing vault. Only callregister_vaultorcreate_vaultwhen the user explicitly asks to add or create one. forget_vaultonly removes a remembered entry, but still requires explicit user intent andconfirm: true. It never deletes the vault folder.- Use
search_resources, tag/property queries, and graph tools to narrow the relevant notes before reading full Markdown. - Read the selected notes before summarizing them. Distinguish note content from your own inference.
- Use connections, orphans, dead ends, and placeholders to explain knowledge-graph structure rather than inferring links from titles alone.
- Treat workspace-relative paths returned by the tools as stable note identifiers for follow-up calls.
- The default server is read-only. Do not ask for or imply write access unless the user explicitly requests a note change.
- When write tools are unavailable, return a proposed Markdown patch in chat instead of editing the vault by another route.
- Before changing an existing note, call
preview_resource_updateand review its diff. Apply the same proposal withupdate_resourceand the returnedexpected_content_sha256only when the requested change authorizes it. - If
update_resourcereportsconflict, read the note again and create a new preview. Never retry with an old hash or overwrite newer content. - For new notes, pass the complete Markdown as
content. Use an explicit workspace-relative.mdpathfor CJK or otherwise non-ASCII filenames.
What ships with it
1 file 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.
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 · 24 lines · 71 tokens per session scan A 01e2e34ae0a0
comet-knowledge is a skill published in the GitHub repository Avery320/codex-vault (0 stars, last pushed 9d ago), licensed MIT. It adds 71 tokens to every session and 512 once invoked, about $0.0004 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…