now live! · v1.0

Agents that learn
from every run.

The open-source contextual learning layer for your agents.

The Spolm Data Pipeline

1
2
3
01 · trace/ capture

Spolm records every run.

Our tracing SDK seamlessly integrates into your agent.

Trace
Analysis
Knowledge Base
spolmPersonalLogse28b300f-406c-45d1…
GmailReadAndReply
Task: Read the latest email and reply with 'chicken' in it
Run ID e28b300f-406c…Duration 7.147s
Steps
11
Status
Complete
OverviewLogsIssuesRubric Evals
Steps
1. reasoning
llmgemini
756msfailure
2. authenticate
authgmail_oauth
2mssuccess
3. reasoning
llmgemini
563mssuccess
4. fetch_email
emailgmail
292mssuccess
5. reasoning
llmgemini
776mssuccess
6. summarize
llmgemini
763mssuccess
7. reasoning
llmgemini
776mssuccess
8. generateReply
llmgemini
1062mssuccess
9. reasoning
llmgemini
927mssuccess
10. sendReply
emailgmail
524mssuccess
11. reasoning
llmgemini
664mssuccess
reasoning
Step 3c2a574b-ca03-44a4-ba72…
failure
INPUT
{
  "goal": "Read the latest email and reply with 'chicken' in it",
  "iteration": 1,
  "context": {
    "lastAction": null,
    "authenticated": false
  }
}
OUTPUT
{
  "action": "authenticate",
  "reasoning": "The agent is not authenticated, so it needs to authenticate first."
}
02 · score/ grade

Every run, graded.

Rubrics score the trace end-to-end. Spolm pinpoints what broke and exactly where.

Trace
Analysis
Knowledge Base
spolmLogsRun Overview
Task: What is quantum computing?
Run ID d96995ef-8ce7-4472-ab42…·Duration 59.88s·Tokens 0
complete
Score
0/10
0 passed · 0 failed
✓ What went right
✕ What went wrong
◐ Root cause summary

03 · learn/ remember

Every mistake, remembered.

Lessons from graded runs feed back as context before the next run starts.

Trace
Analysis
Knowledge Base
Session 1 · Yesterday
Run #182
8 steps · 4.9s · 2 failures
reasoning756ms
reasoning612ms
fetch_email401ms
authenticate4ms
fetch_email292ms
reasoning776ms
generateReply1062ms
sendReply524ms
Lesson learned
Always authenticate before fetch_email
accuracy +24%tool sequence
Knowledge base
246 patterns
retrieved as context
Session 2 · Today
Run #183
4 steps · 1.8s · 0 failures
authenticate3ms
fetch_email280ms
generateReply998ms
sendReply510ms

Built for what agents actually need.

COST PER RUN
cost
Fewer tokens, same results.

Spolm surfaces what your agent already knows from past runs, so it stops re-deriving context from scratch every time. Token costs fall as the knowledge base grows.

BEFOREAFTERfailures drop as memories accumulate
reliability
Failures teach the next run.

Warnings and lessons from failed runs are stored and injected as context before the next run. Your agent learns what to avoid without any code changes.

youragentspolmsdkframework #1framework #2custom agent
integration
Works with your stack.

Seamless integration into LangChain, OpenAI ADK, or custom agents, with more adapters coming soon. No architecture changes, no new infrastructure to manage on your end.

See it on your own agents.