Skip to content

Align skills with unified MCP access and sandbox workflows - #11

Open
xav-db wants to merge 2 commits into
mainfrom
codex/fix-unified-mcp-skills
Open

xav-db wants to merge 2 commits into
mainfrom
codex/fix-unified-mcp-skills

Conversation

@xav-db

@xav-db xav-db commented Sep 18, 2026

Copy link
Copy Markdown
Member

Move query/admin skill dependencies to the unified MCP endpoint and select Cloud workflows by authenticated identity. Agents with sandbox scopes and service credentials no longer get blocked by a mandatory human-only observability step.

One MCP endpoint
├── helix-mcp → human OAuth inspection
├── helix-query-mcp → permitted queries and agent sandbox setup
└── helix-admin-mcp → permitted customer administration

Updates query authoring, translation, optimization, and memory guidance to allow static work when telemetry is unavailable without claiming live verification. Corrects tenant creation to return no application key and distinguishes rejected confirmation bindings from consumed confirmations. Keeps explicit execution, exact payload binding, and no automatic mutation retries.

Adds a CI publication check and six MCP benchmark cases for human discovery, service credentials, sandbox setup, missing tools, tenant creation, and confirmation failures.

Validation: python3 scripts/check-mcp.py, Ruff lint/format, and git diff --check passed. Negative checks rejected a retired endpoint, implicit execution, missing agent setup guidance, and the obsolete tenant-key claim. Benchmark prompts and scoring checklists were added; an automated agent benchmark and authenticated production acceptance were not run.

Companion docs: HelixDB/helix-db#1103

xav-db added a commit to HelixDB/helix-db that referenced this pull request Sep 19, 2026
Correct the MCP docs that send customers to separate admin services and
reject service credentials on the unified endpoint. Document the current
access boundaries and executable tool arguments, and regenerate the
agent documentation indexes.

```text
mcp.helix-db.com/mcp
├── Human OAuth → permitted inspection, query, and admin tools
├── Service credential → granted query and admin tools
└── Agent registration → sandbox setup and scoped sandbox queries
```

Includes read, prepare/execute write, and tenant-creation examples;
clarifies missing telemetry, separate application-key creation, and
confirmation handling. Adds publication checks for retired endpoint/auth
guidance and matching write confirmation examples.

Validation: `npm run check` in docs (page/JSON contracts, generated
indexes, OpenAPI), Prettier on changed scripts, and `git diff --check`
passed. Negative checks rejected a retired endpoint, invalid database
reference, and missing confirmation token. Reviewed against the current
backend tool registration and argument schemas. No Rust runtime changes;
authenticated production acceptance and deployment are outside this PR.

Companion skills PR: HelixDB/skills#11



<!-- greptile_comment -->

<!-- greptile_summary -->

<h2><a
href="https://app.greptile.com/api/retrigger?id=66010966"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://greptile-static-assets.s3.amazonaws.com/badges/RetriggerDark.svg?v=2"><source
media="(prefers-color-scheme: light)"
srcset="https://greptile-static-assets.s3.amazonaws.com/badges/Retrigger.svg?v=2"><img
alt="Retrigger"
src="https://greptile-static-assets.s3.amazonaws.com/badges/Retrigger.svg?v=2"
align="right"></picture></a></h2>

The PR appears safe to merge, with two non-blocking
documentation-validation improvements recommended.

<h3>Summary</h3>

This PR updates the Helix Cloud documentation to describe unified MCP
authentication for human OAuth, service credentials, and agent
registrations. It also adds executable query and administration
workflows, regenerates the LLM-facing documentation indexes, and
introduces publication checks for stale MCP guidance and write
confirmations.

- Replaces references to separate query/admin MCP services with the
unified endpoint.
- Documents identity-specific tool and permission boundaries.
- Adds read, confirmed-write, and tenant-creation examples.
- Adds documentation validation, although administration examples remain
unchecked and write validation is tied to a fixed snippet count.

<details><summary>Diagram</summary>

```mermaid
%%{init: {'theme': 'neutral'}}%%
flowchart TD
  E["mcp.helix-db.com/mcp"]
  H["Human OAuth"] --> E
  S["Service credential"] --> E
  A["Agent registration"] --> E
  E --> HD["Cloud discovery and observability"]
  E --> Q["Authorized query tools"]
  E --> C["Customer administration tools"]
  E --> SB["Ready sandbox setup and scoped queries"]
  H --> HD
  H --> Q
  H --> C
  S --> Q
  S --> C
  A --> SB
```
</details>

<sub>Reviews (1) · Last reviewed commit: ["Format documentation index
generator"](https://github.com/helixdb/helix-db/commit/700e606ac256a77c75f581c1af606155e3dcd998)</sub>

<!-- /greptile_comment -->
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant