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/yashvanthange/supermemory/supermemory-agent-learningnpx skills add YashvantHange/SuperMemory --skill supermemory-agent-learninggit clone --depth 1 https://github.com/YashvantHange/SuperMemoryWhat 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 | $0.00077 | $0.01283 |
| Opus 5 | $0.00039 | $0.00642 |
| Sonnet 5 | $0.00015 | $0.00257 |
| Haiku 4.5 | $0.00008 | $0.00128 |
Grade A, and why
supermemory-agent-learning 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 — 158 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SuperMemory — Agent Learning Layer
Core principle (never violate)
SuperMemory is not full observability. Capture workflow outcomes and distilled lessons only — never full conversation transcripts.
| Capture | When |
|---|---|
| Workflow snapshot | Run start/end |
| Failure / correction / suggestion | Significant events only |
| Distilled lesson | After reflect(event_ids) → validate → process_promotions |
Token budget at retrieval: default max_tokens=800. Store lessons, not histories.
Architecture (closed loop)
retrieve → record_failure → reflect(event_ids) → validate → process_promotions
→ retrieve again → report_outcome
Critical gates: Memory Validator (reject/merge/rewrite/approve) and Promotion Queue (no immediate store without validation).
Project layout
src/supermemory_mcp/ # FastMCP server (13 core + 16 learn.* tools, MCP resources)
packages/uall/ # UALL engine (validator, promotion, retrieval, experiments)
packages/uall_server/ # FastAPI REST
packages/uall_python/ # Python SDK
storage/adapters/ # file (.supermemory/), sqlite, postgres
skills/ # canonical agent skills (this file)
examples/ # MCP configs + agent demos
tests/ # 69 tests incl. real stdio MCP transport
Entry point: packages/uall/service.py (UALLService).
When to use which interface
| Context | Use |
|---|---|
| Claude / Cursor agents | MCP — python -m supermemory_mcp.server |
| Python agent in-process | UALLClient from uall_python |
| Remote / polyglot | REST — python -m uall_server (port 8000) |
| Tests | python -m pytest tests/ -v |
Integration workflow (agent developer)
- [ ] 1. retrieve — policies + lessons before the step (stage-aware)
- [ ] 2. record_failure / record_correction — selective events only
- [ ] 3. reflect(event_ids) — evidence-first, never free-text-only
- [ ] 4. validate → process_promotions — explicit promotion step
- [ ] 5. retrieve again on next run
- [ ] 6. report_outcome — telemetry recalibrates confidence
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 · 158 lines · 77 tokens per session scan A 4c1dc260cca4
supermemory-agent-learning is a skill published in the GitHub repository YashvantHange/SuperMemory (2 stars, last pushed 2mo ago), licensed MIT. It adds 77 tokens to every session and 1,283 once invoked, about $0.0004 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
agent-coordination
Find other agents, list participants in a conversation, send routed messages, attach files, and recover earlier conversation context (history / resolve a fuzzy reference / read a quoted message / read compressed summaries). Use whenever you need to delegate to another agent, address a peer in a multi-agent…
claim-agent-ownership
Orchestrator skill for resolving multi-daemon binding contention. Use when you (the orchestrator) detect an agent.binding.contested sync event indicating two daemons are racing for the same agent — explicitly rebind ownership to a chosen target daemon so subsequent dispatches route deterministically. Implements Gap…
human-approval
Request human approval before performing a SAFETY-CRITICAL, IRREVERSIBLE, or SCOPE-EXPANDING action — submit a structured context (action, scope, risk, consequence) plus options, then STOP the current turn. The platform redispatches the agent after the human decides. NEVER use for routine deliverables (writing docs /…
skill-authoring
Generate Prismer-compliant skill drafts from user intent, documentation URLs, existing code, or service endpoints. Use whenever the user says "make this a skill", "package this workflow", "create a skill for X", or wants to capture a repeatable workflow into a reusable artifact. Outputs a multi-file manifest (SKILL.md…
plugin-dev
Prismer Evolution Plugin 开发指南 — 快速迭代 hook/skill、调试、日志查看、测试、发布全流程.
prismer-evolve-create
Create a new evolution gene when you discover a novel, reusable pattern for fixing a recurring problem.