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/michielhdoteth/squish-memory/squishnpx skills add michielhdoteth/squish-memory --skill squishgit clone --depth 1 https://github.com/michielhdoteth/squish-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/michielhdoteth/squish-memory/squish)<a href="https://agentmods.dev/skills/michielhdoteth/squish-memory/squish"><img src="https://agentmods.dev/badge/skills/michielhdoteth/squish-memory/squish.svg" alt="Measured on agentmods" 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 | $0.00157 | $0.01799 |
| Opus 5 | $0.00078 | $0.00899 |
| Sonnet 5 | $0.00031 | $0.00360 |
| Haiku 4.5 | $0.00016 | $0.00180 |
Grade A, and why
squish 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 3d 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 — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Squish: Memory Runtime for AI Agents
Connect once. Remember everywhere. Squish gives ChatGPT, Claude Code, and every AI agent one shared memory through OAuth. Stop re-explaining your project, context, and decisions to every tool.
How Squish Works
Squish is a memory runtime that:
- Captures durable signal from agent sessions automatically
- Stores memories with embeddings, beliefs, and relationships
- Retrieves relevant context across sessions using hybrid search
- Decays unimportant memories over time using Ebbinghaus curves
- Syncs via OAuth cloud when teams need shared memory
Installation
Option A: Cloud (Recommended - connect to ChatGPT/Claude in 2 minutes)
npx squish-memory
squish cloud login
Then configure your agent to use the Squish MCP server.
Option B: Local (fastest, no cloud dependency)
npm install -g squish-memory && squish install --all
MCP Configuration
For OpenCode
Add to your opencode.json:
"squish": {
"type": "local",
"command": ["squish-mcp"],
"enabled": true
}
For Claude Desktop
{
"mcpServers": {
"squish": {
"command": "npx",
"args": ["-y", "squish-memory"]
}
}
}
For ChatGPT (Cloud)
- Go to ChatGPT Settings > Custom MCP Servers
- Server URL:
https://api.squishplugin.dev/mcp - Auth: Bearer Token (get your key from squishplugin.dev)
Tool Reference
When Squish is connected, these 15 tools are available:
| Tool | Params | What it does |
|---|---|---|
squish_remember |
content, type?, tags? |
Save content, decisions, preferences, or facts. Auto-detects routing (memory vs learning vs belief). |
squish_recall |
query, limit?, project? |
Search memories, learnings, beliefs, and graph entities. Hybrid search. Returns a recall assessment (verdict: confident / qualified / no_reliable_memory). |
squish_forget |
memoryId?, search? |
Delete a single memory by ID (auto-confirm) or bulk by search query (dry-run with preview). |
squish_link |
action, memoryId?, fromId?, toId? |
Manage memory associations. action=find traverses the graph. action=add links two memories. |
squish_context |
project?, actorUser? |
Auto-load relevant context for current project including beliefs, contradictions, and staleness warnings. |
squish_stats |
project? |
Show memory statistics, signal counts, graph status, places, and system health checks. |
squish_inspect |
memoryId |
Show why a memory was stored, its classification, beliefs, place, and graph status. |
squish_skill |
action, ... |
Manage reusable skill documents extracted from memory patterns. |
squish_loadout |
action, ... |
Bind memory assets to agents; manage visibility rules (ACL). |
squish_extract |
action, hoursBack? |
Auto-extract reusable skills from accumulated memories via LLM analysis. |
squish_feedback |
targetType, id, signal |
Reinforce or weaken a recalled item: confirm (correct), used (acted on), or contradict (wrong). Targets: memory, belief, or strategy. |
squish_places |
action, ... |
Query and manage cognitive places (inbox, wip, ref, board, sparks). |
squish_sessions |
action, ... |
Search past agent sessions. Sources: opencode, claude-code, codex, gemini, all. |
squish_tier |
action, ... |
Inspect and manage memory tiers (working / long-term / cold). |
squish_dedup |
action, ... |
Duplicate detection and merge proposals with review + reverse support. |
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.
- 3d ago First seen · 146 lines · 157 tokens per session scan A c68730b1080b
squish is a skill published in the GitHub repository michielhdoteth/squish-memory (11 stars, last pushed 6d ago), licensed MIT. It adds 157 tokens to every session and 1,799 once invoked, about $0.0008 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
docmancer
Work from the same local memory as every other coding agent on this machine. Recall prior decisions, preferences, instructions, and project conventions that Claude Code, Codex, Cursor, and other agents wrote here, with cited sources, fully local. Also searches a separate local technical-documentation index.
memos-local
Persistent local memory for OpenClaw agents. Use when users say: "install memos" "install MemOS" "setup memory" "add memory plugin" "openclaw memory" "memos onboarding" "memory not working" "configure memory" "enable memory" "upgrade MemOS" "update memory plugin".
memoryguard
Local-first MCP memory backend and governance console for coding agents. Auto-organize, quarantine, supersede, and rollback shared memories across multiple agents.
mem9
Persistent cloud memory for OpenClaw agents. Use when users say: "install mem9" "setup memory" "add memory plugin" "openclaw memory" "mem9 onboarding" "memory not working" "import memories" "upload sessions".
mem9
Persistent cloud memory for OpenClaw agents. Use when users say: "install mem9" "setup memory" "add memory plugin" "openclaw memory" "mem9 onboarding" "memory not working" "import memories" "upload sessions".
ourmem
Shared memory that never forgets. Cloud hosted or self-deployed. Collective intelligence for AI agents with Space-based sharing across agents and teams. Use when users say: "install ourmem" / "install omem" "setup memory" / "setup omem" "add memory plugin" "ourmem onboarding" / "omem onboarding" "memory not working"…