peta-core

Peta Core is a self-hosted control-plane/runtime for MCP (Model Context Protocol): it proxies MCP calls through a gateway, manages downstream server lifecycles, encrypts/stores credentials in a vault, enforces RBAC/ABAC policies (optionally with human-in-the-loop approvals), and writes structured audit logs. It also provides an embedded OAuth 2.0 authorization server and can expose anonymous/public access for selected servers.

Evaluated Mar 30, 2026 (21d ago)
Homepage ↗ Repo ↗ Infrastructure ai-ml mcp gateway security authentication vault audit policy-engine typescript oauth2 rbac abac self-hosted
⚙ Agent Friendliness
64
/ 100
Can an agent use this?
🔒 Security
77
/ 100
Is it safe for agents?
⚡ Reliability
40
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
85
Documentation
65
Error Messages
0
Auth Simplicity
70
Rate Limits
70

🔒 Security

TLS Enforcement
80
Auth Strength
85
Scope Granularity
75
Dep. Hygiene
45
Secret Handling
90

README claims credentials are encrypted at rest (PBKDF2 + AES-GCM) and injected server-side at execution time, and that secrets are excluded from logs. It also describes RBAC/ABAC with optional human-in-the-loop approvals and audit logging. Dependency hygiene score is estimated due to lack of explicit vulnerability/SBOM/CVE remediation data in the provided manifest/README excerpt.

⚡ Reliability

Uptime/SLA
0
Version Stability
40
Breaking Changes
50
Error Recovery
70
AF Security Reliability

Best When

You need a production-grade MCP control plane with authentication/authorization, credential vaulting, auditability, and managed downstream connectivity.

Avoid When

You only require local/dev experimentation where the additional operational complexity (OAuth server, DB, supervision, vault keys) isn’t justified.

Use Cases

  • Run MCP in production with a centralized gateway/proxy for multiple downstream MCP servers
  • Centralize and enforce per-user, per-tool authorization (RBAC/ABAC) before tool execution
  • Store and inject downstream credentials securely at execution time
  • Provide audit trails for every tool call (without logging secrets)
  • Implement approval workflows for high-risk tool actions
  • Expose REST endpoints as MCP tools via a REST-to-MCP adapter

Not For

  • Building a simple single-purpose MCP server without gateway/vault/policy needs
  • Environments that cannot run/operate a self-hosted backend component (deployment, DB, keys, etc.)
  • Use as a generic credential store without the surrounding policy/audit gateway context

Interface

REST API
Yes
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
Yes
Webhooks
Yes

Authentication

Methods: OAuth 2.0 authorization server (Authorization Code + PKCE, refresh tokens) Dynamic client registration Token introspection Token revocation Anonymous public access mode for selected public servers (separate /mcp/public endpoint)
OAuth: Yes Scopes: Yes

Auth is described as including an embedded OAuth2 server and capability/policy-based filtering. Exact scope model and enforcement details are not included in the provided README excerpt.

Pricing

Free tier: No
Requires CC: No

Open-source/self-hosted component (no SaaS pricing described).

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Documented

Known Gotchas

  • Approval-gated or cached results may change tool execution semantics (tool might not run immediately if human-in-the-loop is required).
  • Public/anonymous access is available only for selected servers on a separate endpoint; ensure the agent targets the correct route.
  • Downstream retries (up to two reconnect/retry attempts) may still cause non-idempotent side effects depending on the downstream tool.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for peta-core.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered