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 skills add ozankasikci/global-agent-memory --skill global-memorygit clone --depth 1 https://github.com/ozankasikci/global-agent-memoryWrote 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/ozankasikci/global-agent-memory/global-memory)<a href="https://agentmods.dev/skills/ozankasikci/global-agent-memory/global-memory"><img src="https://agentmods.dev/badge/skills/ozankasikci/global-agent-memory/global-memory/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/skills/ozankasikci/global-agent-memory/global-memory"><img src="https://agentmods.dev/badge/skills/ozankasikci/global-agent-memory/global-memory.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 | $0.00074 | $0.01071 |
| Opus 5.5 | $0.00030 | $0.00428 |
| Sonnet 5.5 | $0.00015 | $0.00214 |
| Haiku 4.5 | $0.00007 | $0.00107 |
Grade A, and why
global-memory 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 1mo 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Global Agent Memory
Skill version: 1.2.0
MCP contract version: v1
Use the canonical MCP capability names. Do not assume a client-specific server prefix. Read references/contract-v1.md only when exact capability selection or arguments are unclear.
Before substantial work
- Derive the project from the current working directory. Pass
working_directory; addprojectonly when the user supplied or confirmed it. - Call
memory_contextwith the concrete task before broad searches. Keepcross_project=falseunless the user explicitly requests comparison or wider scope. - Use
memory_searchfor a prior decision, exact ID, symbol, error string, convention, or known solution that the context bundle did not answer. - If retrieval returns
protected_memory_may_be_relevant, do not guess at hidden metadata. Callmemory_access_requestwith the concrete purpose, query, resolved project, least required permission, and shortest useful duration. Tell the user a request is waiting in the dashboard, where they choose the exact memories and may reduce the requested permission or duration. Pollmemory_access_status, then pass an approvedaccess_grantonly to the operation it was requested for. - Use
memory_getbefore relying on, citing, or changing one result. Treat status and applicability labels as constraints, not decoration. - When memory materially informs the work, cite its memory ID and Vault-relative path in the response or handoff.
Do not turn retrieval into ceremony for trivial work. If the available context is sufficient and no durable history could change the answer, proceed.
Open the dashboard
When the user explicitly asks to open, show, or launch the Global Agent Memory dashboard, call memory_dashboard_open
with open_browser=true. Do not invent a URL or substitute Obsidian. The tool issues a short-lived authenticated local
session and starts the localhost dashboard server on demand when necessary.
Agents never approve, deny, or revoke access requests. Sealed memories are owner-only and cannot be retrieved through MCP. Never store credentials or secrets as memory.
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.
- 1mo ago Changed · +2 lines 959ebeb48410
- 1mo ago First seen · 77 lines · 74 tokens per session scan A 895c15046e04
global-memory is a skill published in the GitHub repository ozankasikci/global-agent-memory (14 stars, last pushed 1mo ago), licensed MIT. It adds 74 tokens to every session and 1,071 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
memex-best-practices
Zettelkasten best practices for building a high-quality knowledge graph.
memex-agentic-memory
A-MEM-inspired agentic memory workflow for structured knowledge capture.
memex-organize
Periodic maintenance of the Zettelkasten card network.
memex-retro
Save insights from completed tasks to Zettelkasten memory.
memex-recall
Load prior knowledge from Zettelkasten memory when the task likely benefits from past context.
memex-sync
Sync Zettelkasten cards across devices via git.