docs(examples): agents-of-all-shapes — any framework → one Tangle Intelligence pipe (no sandbox)#72
Merged
Merged
Conversation
…elligence pipe (no sandbox) A QA-verified showcase that Tangle Intelligence works with ANY agent, not just our sandbox. Tangle runtime, OpenAI-compatible router (tcloud/OpenRouter), Mastra, Claude Agent SDK, and a Python agno agent all converge on the same canonical OTel GenAI spans → fromOtelSpans → analyzeRuns → InsightReport, computed in-process (no sandbox, no deploy, no server). - shared/intelligence.ts: the one contract — toInsightReport(spans) + shipToTangleOtlp(spans) over standard gen_ai.* OTel attributes + score - shapes.ts: 4 TS shapes as deterministic batches (CI-verifiable, no key) with the exact live wiring documented per framework - python-agno/: runnable agno → OTLP script - run.ts: merges all shapes → fleet InsightReport + per-shape breakdown - tests/agents-of-all-shapes.test.ts: proves each shape → real InsightReport + cross-framework fleet aggregation + real cost from gen_ai.usage.cost_usd 377/377, typecheck + biome clean.
…mat fix from #70 was lost in merge)
tangletools
pushed a commit
that referenced
this pull request
May 28, 2026
…gents-of-all-shapes showcase (#72)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
A QA-verified showcase proving Tangle Intelligence works with any agent, not just our sandbox. Five shapes — Tangle runtime, OpenAI-compatible router (tcloud / OpenRouter), Mastra, Claude Agent SDK, Python agno — all converge on the same canonical OTel GenAI spans and the same in-process engine:
agent → OTel GenAI spans → fromOtelSpans → analyzeRuns → InsightReportNo sandbox, no deploy, no ingest server — the analysis runs in-process.
What's here
shared/intelligence.ts— the entire integration surface:toInsightReport(spans)(in-process QA path) +shipToTangleOtlp(spans)(optional hosted path), over standardgen_ai.*OTel attributes +score.shapes.ts— 4 TypeScript shapes as deterministic batches (verifiable in CI with no LLM key) with the exact live wiring documented per framework.python-agno/agno_to_intelligence.py— runnable agno → OTLP/HTTP script.run.ts— merges all shapes → fleetInsightReport+ per-shape breakdown.tests/agents-of-all-shapes.test.ts— proves each shape → a realInsightReport, cross-framework fleet aggregation, and real cost derived fromgen_ai.usage.cost_usd.Why
The integration point is the OTel wire, not the Tangle SDK or sandbox. Any team with agent traces gets the full decision packet without adopting our execution stack — exactly the pilot customer's "we won't use your sandbox but will use the libraries" path.
Test plan
pnpm test -- tests/agents-of-all-shapes.test.ts(3/3); full suite 377/377pnpm typecheckcleanbiome check src tests examplesclean