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/metraton/gaia/memorynpx skills add metraton/gaia --skill memorygit clone --depth 1 https://github.com/metraton/gaiaWhat 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 | $0.00047 | $0.01817 |
| Opus 5 | $0.00023 | $0.00908 |
| Sonnet 5 | $0.00009 | $0.00363 |
| Haiku 4.5 | $0.00005 | $0.00182 |
Grade A, and why
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 2d 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 — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Memory
Memory is the curation technique that turns selected experience into continuity: durable knowledge kept small, live work kept visible, and raw operational history left on the automatic event floors where it belongs.
The three floors
| Floor | Purpose | Lifecycle |
|---|---|---|
| Events | Commands, dispatches, session events, and other operational facts | Automatic and short-lived |
| Episodes | Searchable agent-turn outcomes and anomalies | Automatic, retained for diagnosis |
| Curated memory | User-governed knowledge and work that must affect future decisions | Deliberate and long-lived |
Events and episodes are evidence, not durable truth.
Think in roles, not storage vocabulary
Every curated item serves one human-facing role:
- Durable knowledge: stable facts, accepted decisions, useful dead ends, user preferences, and meaningful milestones.
- Live thread: one actionable concern that must reappear in a later session.
- Historical log: append-only context useful for audit but not reinjection.
The internal class, status, type, slug, and link enums implement these
roles. Consult reference.md only when materializing or debugging them.
Other home first
Before saving, ask: does this already have a canonical home? Work in flight belongs in a brief, plan, or task; domain state in project context or the owning system; raw execution detail in events, episodes, or the transcript.
Do not copy a fact into memory merely because it matters. A second source of truth becomes stale; a durable reference to the canonical object is enough.
The one-line initiative test
- Observed Gaia fail or rub in use:
initiative=gaia_system, host-scoped bygaia/store/writer.py::apply_host_scope; a project anchor is refused. - Decided to build or change Gaia: project-scoped
initiative=gaia.
Choose from how the fact was produced, not the current directory. The writer
enforces the destination after that choice; reference.md owns its mechanics.
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.
- 2d ago First seen · 151 lines · 0 tokens per session scan A 59b6b0c79f6a
memory is a skill published in the GitHub repository metraton/gaia (3 stars, last pushed 5d ago), licensed MIT. It adds 47 tokens to every session and 1,817 once invoked, about $0.0002 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
python-feature-lifecycle
Guidance for package and feature lifecycle in the Agent Framework Python codebase, including stage meanings, feature-stage decorators, feature enums, and how to move APIs from one stage to the next.
python-development
Coding standards, conventions, and patterns for developing Python code in the Agent Framework repository. Use this when writing or modifying Python source files in the python/ directory.
foundry-config-setup
Resolve missing setup caused by a hardcoded Foundry project endpoint or model in a sample. Use when a sample fails because it uses a placeholder/hardcoded projectendpoint (for example "https://your-project.services.ai.azure.com") or a hardcoded model instead of reading them from the environment.
reflect
Review recent work, find repeated workflow patterns, and suggest reusable skills, agents, commands, config changes, or playbooks. Use when the user asks to learn from past sessions, improve recurring workflows, or identify what should be turned into reusable agent instructions.
codemap
Generate comprehensive hierarchical codemaps for UNFAMILIAR repositories. Expensive operation - only use when explicitly asked for codebase documentation or initial repository mapping.
length-converter
Convert between common length units (miles, km, feet, meters) using a multiplication factor.