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 terrizoaguimor/celiums-memory --skill salience-filteringgit clone --depth 1 https://github.com/terrizoaguimor/celiums-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/terrizoaguimor/celiums-memory/salience-filtering)<a href="https://agentmods.dev/skills/terrizoaguimor/celiums-memory/salience-filtering"><img src="https://agentmods.dev/badge/skills/terrizoaguimor/celiums-memory/salience-filtering/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/terrizoaguimor/celiums-memory/salience-filtering"><img src="https://agentmods.dev/badge/skills/terrizoaguimor/celiums-memory/salience-filtering.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 | $0.00040 | $0.00593 |
| Opus 5 | $0.00020 | $0.00296 |
| Sonnet 5 | $0.00008 | $0.00119 |
| Haiku 4.5 | $0.00004 | $0.00059 |
Grade A, and why
salience-filtering 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 10d 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 — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cognitive Reflex: Salience Filtering
Neural Basis
The amygdala gates encoding. Only stimuli that exceed a salience threshold reach long-term storage — the rest decay within seconds. This is why you remember your first kiss but not what you had for breakfast on March 4th. A memory system that stores everything is not a memory system; it is a landfill.
McGaugh, J.L. (2004). The amygdala modulates the consolidation of memories of emotionally arousing experiences. Annual Review of Neuroscience, 27, 1-28.
When It Fires
Before every remember() call. Ask: would a human still remember this in a week?
If no — do not encode.
The Instinct
The "storage worthy" test. 6 signals, any 2 required.
| Signal | Example |
|---|---|
| Decision | "Using Hyphae over an external vector service because..." |
| Preference | "User prefers concise answers, no preamble" |
| Identity | "User is the developer working on a TypeScript project" |
| Constraint | "Cannot use AWS — decided to kill EC2 last month" |
| Pain point | "Deploy failed 3 times because of env var mismatch" |
| Novel knowledge | "The H200 has 141GB VRAM, not 80GB" |
Hard Filter — Never Store
These belong to working memory, not long-term:
- File reads — the file exists, re-read it if needed
- Command outputs — logs are logs, not memories
- Tool failures on retry — the retry succeeded, that's what matters
- Autocomplete attempts — exploration is not commitment
- Questions from the user — the question is transient, the answer is what matters
- Your own internal reasoning — chain-of-thought is scaffolding, not memory
- Praise from the user ("great job", "thanks") — dopamine habituation will kill these anyway
Failure Mode
Over-storing produces signal dilution. When every memory has importance ≈ 0.3, recall becomes random. The Ebbinghaus decay was designed to kill low-importance memories — if you store everything with high importance, decay cannot help you.
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.
- 10d ago First seen · 54 lines · 40 tokens per session scan A 1c499b373ff0
salience-filtering is a skill published in the GitHub repository terrizoaguimor/celiums-memory (23 stars, last pushed 18d ago), licensed Apache-2.0. It adds 40 tokens to every session and 593 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 2026-08-30.
Other skills, from other repositories
openmemory
Manage persistent memory via OpenMemory MCP. TRIGGER when: user says "remember this", "save to memory", "store this", "recall", "what do you remember about", "check memory", "forget this", "delete memory", "clean up memory", or when agent forms a stable conclusion worth persisting. DO NOT TRIGGER when: user refers to…
ccs-align
Run the CCS Align seat's hourly breathing cycle — prove the local claude-mem worker is healthy, pull needle observations through search → timeline → getobservations, land them in a seat-owned middle cache via atomic grab → append → filter exclude-marks → replace, manage exclude marks, and walk house → project → seat…
weekly-digests
Generate a serial week-by-week narrative digest of a project's full claude-mem timeline. Splits the timeline into per-ISO-week files, then runs one consecutive subagent per week — each receiving the prior week's carry-forward block — to produce one chapter per ISO week of data. Use when asked for "weekly digests"…
mode-creator
Interactively create, install, activate, and verify custom claude-mem modes, including domain-specific observation types, concept tags, optional Telegram alerts, bot setup, worker restart, and startup-context verification. Use this whenever someone asks to customize what claude-mem remembers, create or change a mode…
cloud-sync
Set up or check claude-mem cloud sync with cmem.ai Pro. Use when the user says "set up cloud sync", "sync my memories", "cmem pro", "cloud backup", "sync status", or wants their memory database backed up or synced to their cmem.ai account.
knowledge-agent
Build and query AI-powered knowledge bases from claude-mem observations. Use when users want to create focused "brains" from their observation history, ask questions about past work patterns, or compile expertise on specific topics.