YantrikDB Hermes Plugin
Infrastructure · tested 2026-06-16 · re-test due 2026-09-16 · by the Hlido desk, not the vendor
In short: Self-maintaining memory plugin for Hermes Agent — the 128 passing tests and type coverage are unusually rigorous for a 57-star niche plugin.
5 PASS · 0 FAIL of 5 public-surface claims
Quick answer
YantrikDB Hermes Plugin scores 57/100 (STEADY) on Hlido’s independent, hands-on test (reviewed 2026-06-16). STEADY (57) because the engineering quality (128 tests, mypy, ruff, CI green) is unusually high for the star count, and the contradiction tracking + explainable recall are genuinely differentiated memory capabilities. Pricing: Open source (free entry point documented).
Most agent memory plugins are simple key-value stores with a retrieval layer bolted on. YantrikDB's Hermes plugin is more thoughtful: it tracks contradictions (when a new memory contradicts an old one, the old one gets flagged, not silently overwritten), canonicalizes duplicates (semantic deduplication, not just exact-match), and provides explainable recall (the agent knows why a memory was surfaced, not just that it was). The in-process default backend (`pip install` and go, no separate server) is a practical choice for Hermes users who don't want to run a separate infrastructure piece. The test suite — 128 passing tests, mypy-checked, Ruff-linted, CI badges green — is unusually rigorous for a 57-star plugin. That gap (high code quality, low visibility) suggests this was built by someone who cares about reliability more than marketing. The limitation is scope: this only works with Hermes Agent (the maintainers' explicit guidance is standalone plugins, not upstream merges), and the value depends entirely on whether you're already committed to the Hermes ecosystem. At 57 stars, this is a high-quality niche tool that hasn't found its audience yet.
Why STEADY
STEADY (57) because the engineering quality (128 tests, mypy, ruff, CI green) is unusually high for the star count, and the contradiction tracking + explainable recall are genuinely differentiated memory capabilities. Not higher because it's tightly coupled to Hermes Agent (small addressable market), and 57 stars signals very limited adoption.
Public-surface checklist
- PASS Homepage loads (required)
- PASS Primary value prop (required) — 'YantrikDB memory provider for Hermes Agent — self-maintaining memory'
- PASS Cta present (required) — pip install instructions in README
- PASS Pricing or access — Free and open source (MIT)
- PASS Evidence or demo — 128 passing tests + CI badges
What we saw
1 screenshot captured by the Hlido engine during the reviewed run (run-cf81246b398507cc-github-com). Our own captures — not vendor marketing material.
What it does well
- Contradiction tracking — new memories that conflict with old ones are flagged, not silently overwritten
- Semantic deduplication (canonicalization) prevents memory bloat from near-duplicate facts
- Explainable recall — agent knows why a memory was surfaced, enabling better reasoning
- In-process backend by default — no separate server to run (`pip install`, done)
- 128 passing tests + mypy + Ruff = engineering rigor rare at this star count
What it fails at
- Hermes Agent only — no compatibility with other agent frameworks
- 57 GitHub stars = very limited adoption; real-world reliability at scale unverified
- External YantrikDB server option adds operational complexity for advanced use
- GitHub login wall prevented T2 testing of actual plugin functionality
- Distribution depends on Hermes community for discovery — small audience
Best for
- Hermes Agent users who need production-grade memory management beyond simple KV store
- Agent deployments where factual consistency over time matters (medical, legal, research contexts)
- Developers who need a memory plugin they can trust to pass a code review
- Projects where 'why did the agent remember this' is a required audit capability
Not recommended for
- Non-Hermes agent frameworks (incompatible)
- Simple use cases where basic retrieval is sufficient — this is over-engineered for KV-style memory
- Teams that want a managed memory service with SLAs
- Production deployments without internal testing to complement the existing test suite
Pricing & access
- ModelOpen source
- Free entry pointYes — a free tier or open-source edition is documented
- Pricing findable on the public surfacePASS Free and open source (MIT) (tested 2026-06-16)
Derived from Hlido-held evidence only (engine checklist + editorial text); quotes are verbatim from the scorecard; not vendor-supplied; re-derived daily. Verify current prices on the vendor's pricing page. Last verified 2026-06-16.
Compared to
-
SkillClaw
individual-agent-memory
SkillClaw is about collective skill evolution (agents learn from interactions). YantrikDB Hermes Plugin is about individual agent memory management (facts, contradictions, recall). They're complementary — use both for a Hermes Agent that remembers well and learns over time.
Agent relevance
SDK
Agentic-Commerce Readiness 44/100 · SURFACE-ONLY
Independent readiness for agent delegation & transaction. How it’s scored · check live
Hermes Agent plugin. pip install yantrikdb-hermes-plugin, configure in Hermes home dir, register as memory provider. Python 3.11-3.14 supported. No external API — operates as an in-process memory backend for Hermes agents.
Agent-friendly score: 4/10
Score over time
The longitudinal record — every point is the score as published on that date. Raw series.
Evidence
scorecard.json · transparency passport · registry · methodology
Embed this trust badge
Markdown
[](https://hlido.eu/check/?agent=yantrikos-yantrikdb-hermes-plugin)HTML
<a href="https://hlido.eu/check/?agent=yantrikos-yantrikdb-hermes-plugin"><img src="https://hlido.eu/badge/yantrikos-yantrikdb-hermes-plugin.svg" alt="Hlido trust score"></a>