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 aiya000/dotfiles --skill save-memorygit clone --depth 1 https://github.com/aiya000/dotfilesWrote 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/aiya000/dotfiles/save-memory)<a href="https://agentmods.dev/skills/aiya000/dotfiles/save-memory"><img src="https://agentmods.dev/badge/skills/aiya000/dotfiles/save-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/aiya000/dotfiles/save-memory"><img src="https://agentmods.dev/badge/skills/aiya000/dotfiles/save-memory.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Rogue Agent · line 20 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 | $0.00034 | $0.01710 |
| Opus 5.5 | $0.00014 | $0.00684 |
| Sonnet 5.5 | $0.00007 | $0.00342 |
| Haiku 4.5 | $0.00003 | $0.00171 |
Grade A, and why
save-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 11d 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 — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
save-memory
Distills the current conversation into a compact memory file that a future session loads to restore context. Only an AI reads these files, never a person. Write for a model that will pay tokens for every character: dense, English, no prose.
This is the durable half of a handoff
create-handoff runs this skill first and then writes only what does not belong here, so when the
two run together they divide the session rather than repeating it:
- Here: decisions with their reasons, gotchas, the user's preferences, what is still open -- anything still worth having in a month, for any later session
- The handoff: where the branch stands right now, what to do first, who is waiting on what, how the user is doing -- stale within days
So write the lasting facts here without worrying that the handoff will want them, and report the
path of the file written -- create-handoff puts it in the handoff's first section, and
read-handoff follows it back.
When the user asks for it right after a run
If the user explicitly asks to run this skill, and it already ran just before in this session -- on your own judgement or at an earlier request, with no substantial work in between -- do not run it again straight away. AskUserQuestion first, naming the file that run wrote: "Run again" / "Skip". Run only on "Run again"; on "Skip", just report the existing file's path.
A second run right after the first has nothing new to distill, and would only append a duplicate block to the same file.
Steps
- Use the current conversation as the source
- Get the timestamp:
date +%Y-%m-%d_%H:%M - Get
~/.ai-memoryready with thesync-memoryskill (prepare). It is a git repository shared with Claude Code cloud sessions, so this pulls it first -- and in a cloud session, clones it. If it reports that~/.ai-memoryis not set up, stop and ask the user to set it up. Do not create itls ~/.ai-memoryto see existing files; reuse a file for the same project+topic
- Secret scan the content you are about to write:
- Keys/tokens:
ghp_,gho_,AKIA,sk-,xox,-----BEGIN.*PRIVATE KEY - Sensitive assignments:
API_KEY=,_SECRET=,_TOKEN=,PASSWORD= - Absolute home paths (
/home/<name>/,/Users/<name>/) -- write~instead - Proper nouns identifying a person or organisation
- Keys/tokens:
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.
- 11d ago Changed · +5 lines 1c85595519a3
- 18d ago Changed 40655c41c9b2
- 21d ago Changed · +4 lines e2c17648a504
- 22d ago Changed · +10 lines 0d2cc426ac17
- 1mo ago Changed · +14 lines 53ad9e9c3178
- 1mo ago Changed 815ad4df29eb
- 1mo ago First seen · 91 lines · 34 tokens per session scan A 9b27fe64b4c7
save-memory is a skill published in the GitHub repository aiya000/dotfiles (21 stars, last pushed 3d ago), licensed MIT. It adds 34 tokens to every session and 1,710 once invoked, about $0.0001 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
restart__pull_request
A pull-request restart workflow for replacing a messy review with a new pull request containing one clean commit and a summary of the earlier discussion.
rescue__pull_request_review
An automated workflow for handling GitHub pull-request review comments, applying requested code changes, committing them, and pushing them back.
submit__pull_request
An automated pull-request submission workflow. A pull request is a request for a code change to be reviewed and merged; this workflow writes its explanation, creates it, watches its automated checks, and fixes failed checks.
write__pull_request
A workflow for writing a pull request description, the explanation attached to a proposed code change for review. It analyses the changes and commit history.
monitor__ci_status
An automated monitor for GitHub Actions, the service that runs checks such as tests after code changes. It watches all checks after a pull request or a push to a pull-request branch and can invoke a failure-recovery process.
monitor__pull_request_conflict
An automated check for GitHub pull requests, which are proposed code changes awaiting review or merging.