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/owrede/vault-memory/installnpx skills add owrede/vault-memory --skill installgit clone --depth 1 https://github.com/owrede/vault-memoryWhat 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 | $0.00132 | $0.02740 |
| Opus 5 | $0.00066 | $0.01370 |
| Sonnet 5 | $0.00026 | $0.00548 |
| Haiku 4.5 | $0.00013 | $0.00274 |
Grade D, and why
install scanned grade D with 3 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash - && sudo apt-get install -y nodejs Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -fsSL https://ollama.com/install.sh | sh && (systemctl --user start ollama || ollama serve &) Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **macOS:** Homebrew. If missing: `/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"` How it starts
The opening of the file, as written. The whole thing — 240 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/vmem:install — Guided vault-memory setup
You (the agent) drive this install interactively. Do NOT blindly run a script:
ask the two decisions below first, explain the trade-offs so the user can
choose, then install everything required for the chosen path. Use
AskUserQuestion for the decisions and Bash for the steps. Every step is
idempotent — re-running on a complete setup should report "already done".
The end state of a successful install:
vault-memoryCLI / MCP server installed and onPATH- All dependencies for the chosen engine present (see per-engine sections)
- The target vault(s) registered in
~/.vault-memory/config.toml - Initial index built
.mcp.jsonwritten into each vault root so an MCP-aware client auto-spawns the server- The Obsidian plugin installed into
.obsidian/plugins/vault-memory/(when a vault has.obsidian/) - MCP smoketest passes
Decision 1 — Which retrieval engine?
Ask the user with AskUserQuestion. Present BOTH options with this comparison so
they can make an educated choice:
| Ollama (vector / embeddings) | ContextFit (CPU-only) | |
|---|---|---|
| How it retrieves | Neural embeddings + sqlite-vec ANN + BM25, RRF-fused (semantic search) | Token-native BM25 + Semantic-IDs (lexical/structural) |
| Hardware | Practically needs a GPU (or is slow on CPU); a model stays resident (~1.1 GB for bge-m3) | CPU-only, no GPU, no model (~41 MB deps) |
| Extra dependency | Ollama + a pulled embedding model | Python + contextfit (via pipx) |
| Strengths | Best semantic recall; paraphrase/cross-lingual matching | Fast ingest + query, tiny footprint, fully local, no model download |
| Ideal for | A workstation/laptop with a GPU or spare RAM | Resource-limited / non-GPU hosts (e.g. a Synology NAS), privacy-strict setups |
| Tradeoff | Heavy; GPU/RAM pressure | Lexical, not vector-semantic — exact/structural matches over fuzzy paraphrase |
Both build the same SQLite content layer, so graph/sections/frontmatter/stats tools work either way. The choice only changes the search engine and its dependencies. The engine is per-vault — different vaults can use different engines, and it can be changed later by re-indexing.
What ships with it
2 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.
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 · 240 lines · 0 tokens per session scan D 53d33b472753
install is a skill published in the GitHub repository owrede/vault-memory (0 stars, last pushed 23d ago), licensed MIT. It adds 132 tokens to every session and 2,740 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it D with 3 findings (asks for root, downloads and executes remote code, makes network calls). 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…