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/linxule/memex-plugin/recallnpx skills add linxule/memex-plugin --skill recallgit clone --depth 1 https://github.com/linxule/memex-pluginWhat 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 | $0.00112 | $0.02795 |
| Opus 5 | $0.00056 | $0.01398 |
| Sonnet 5 | $0.00022 | $0.00559 |
| Haiku 4.5 | $0.00011 | $0.00280 |
Grade A, and why
recall 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 — 279 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Recall: Retrieving Session Memory
Context
Project: !basename $(git remote get-url origin 2>/dev/null | sed 's/\.git$//' | xargs basename 2>/dev/null) 2>/dev/null || basename $(pwd)
Vault: !sqlite3 "$(memex path --index 2>/dev/null)" "SELECT (SELECT COUNT(*) FROM fts_content) || ' documents indexed'" 2>/dev/null || echo "(index unavailable)"
Step 0: Classify the Query
Before doing anything, classify the user's question into one of four modes:
TEMPORAL — date-based browsing
Triggers: "yesterday", "last week", "today", "what did I do on Monday", "show me recent work", "last 3 days", "this week's sessions", any date reference without topic keywords.
Action: Go to → Temporal Recall
KEYWORD — topic/decision lookup
Triggers: "why did we...", "find the memo about...", "what was the decision on...", "remind me about the retry pattern", any question with specific technical terms or project names.
Action: Go to → Keyword Recall
DEEP — cross-project synthesis
Triggers: "what patterns do we use across...", "how has our approach to X evolved...", "compare how we handle X in different projects", questions spanning multiple sessions or projects, complex why/how questions needing synthesis.
Action: Go to → Deep Recall
LOAD — direct file retrieval
Triggers: "load the X topic", "pull up project Y", "show me the memo about Z", explicit topic or memo names, "what does the X topic say".
Action: Go to → Load Recall
If mixed (date + topic, e.g., "what auth work did I do last week"): Start with TEMPORAL to narrow the date range, then scan the results for the topic.
Temporal Recall
Run the temporal scanner to browse sessions and memos by date:
memex timeline "<date-expression>"
With project filter:
memex timeline "<date-expression>" --project=<name>
Filter by type:
memex timeline "<date-expression>" --type=memo
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 · 279 lines · 112 tokens per session scan A a95fc5ca32eb
recall is a skill published in the GitHub repository linxule/memex-plugin (7 stars, last pushed 5d ago), licensed MIT. It adds 112 tokens to every session and 2,795 once invoked, about $0.0006 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
orra-init
Initialize orra in the current repo. Runs orrasetup, installs all stock directives, and adds vault-bridge when memento-vault is available. Use when Vic says "init orra", "orra setup", "set up orra here", or invokes /orra-init.
memento
Capture durable knowledge from the current work session into Memento Vault. Use when the user asks to remember, save, capture, or record decisions, discoveries, bug fixes, or reusable patterns.
inception
Run or preview Memento Vault consolidation. Use when the user asks to consolidate, find patterns, run inception, preview clusters, or synthesize cross-session themes.
continue-work
Use when the user says "continue", "pick up where I left off", "what was I working on", or similar. Also use when starting a session and memory has pending work.
concierge
Search Memento Vault for day-to-day recall, past decisions, discoveries, implementation history, and prior session context. Use when the user asks "do you remember", "have we seen", "what did we decide", "what did we learn", "where was this implemented", "find prior context", "check memory", "search the vault", or any…
preserve
Archive artifact bundles in Memento Vault. Use when the user says preserve, archive, or save generated artifacts, or wants evidence bundles kept intact instead of turned into atomic notes.