Commit Graph

7 Commits

Author SHA1 Message Date
Mathias
5950ef5f0f feat(mcpclient): fail-fast on empty bearer token
All checks were successful
CI / Lint / Test / Vet (push) Successful in 10s
CI / Mirror to GitHub (push) Successful in 4s
mcpclient.New previously accepted an empty token and silently omitted
the Authorization header at request time. When the env var sourcing
the token was missing from a Kubernetes Secret (envFrom doesn't warn
on missing keys), this surfaced as an opaque 401 from the upstream
MCP server with no log trail — see hyperguild #13 and brain entry
"mcpclient-empty-token-silent-401-envfrom-missing-key".

mcpclient.New now returns ErrTokenRequired when token is empty.
The routing pod's project_create init checks the error and exits
with a clear message pointing at routing-secrets, turning a runtime
401 storm into a startup crashloop the operator can fix immediately.

Tests pass a dummy "test" token (httptest servers don't enforce
bearer auth, so any non-empty value works). Added a regression
test asserting empty-token construction returns ErrTokenRequired.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-18 16:28:09 +02:00
Mathias
a220fcaf2b feat(routing): create GitHub destination repo before configuring push-mirror
All checks were successful
CI / Lint / Test / Vet (push) Successful in 11s
CI / Mirror to GitHub (push) Has been skipped
Gitea's push-mirror cannot push to a non-existent remote — it just
runs 'git push' against whatever URL it's given. So a project_create
flow that only configures the mirror leaves the GitHub side as an
unfulfillable URL.

New internal/githubclient package: single-purpose client that POSTs
/user/repos to create an empty private repo (auto_init=false so the
first mirror push doesn't conflict with a generated README). Treats
422 'name already exists' as idempotent success via ErrAlreadyExists.
401/403 are surfaced as 'PAT missing repo scope or invalid' so the
operator sees the real cause instead of a vague upstream error.

Skill wiring:
- New stepCreateGitHub between stepCreateRepo and stepMirror in the
  orchestrator.
- Skipped entirely when Config.GitHub is nil (degraded mode — the
  routing pod runs without GITHUB_PAT, mirror config still lands,
  but the actual sync to github fails until the repo exists).
- cmd/routing/main.go constructs githubclient.New(GitHubPAT) only
  when the PAT is set; the skill receives nil otherwise.

Tests:
- happy path: fake github 201 + assertions that the 'reached' array
  is [create_repo, create_github_repo, mirror, infra_commit, issue].
- github 422 already-exists: idempotent, all gitea steps still run.
- github 401: returns failed_step=create_github_repo, no mirror or
  later steps.
- degraded mode (Config.GitHub nil): reached omits create_github_repo,
  rest of the flow runs unchanged.

Updated existing tests to read [skill, gh] from newSkill instead of
just skill, and adjusted reached-array expectations to include the
new step.

Tracks #10.
2026-05-18 13:42:03 +02:00
Mathias
3b79311fdd feat(routing): project_create MCP tool — gitea-first new-project pipeline (#10)
All checks were successful
CI / Lint / Test / Vet (push) Successful in 12s
CI / Mirror to GitHub (push) Successful in 4s
Adds the project_create tool to the routing pod that automates the
"new project" bootstrap end-to-end from claude.ai. Gitea-first
architecture: GitHub receives the repo only via push-mirror, never
via a direct GitHub API call from this server.

Four sequential calls to the gitea-mcp server (configured via
GITEA_MCP_URL):

  1. create_project_from_template — Gitea repo from
     template-go-{agent,web} per the 'stack' arg
  2. repo_mirror_push (action=add) — push-mirror to
     github.com/<GITHUB_OWNER>/<name>.git, interval 8h, sync_on_commit
  3. file_write_branch — k3s/staging/<name>/namespace.yaml committed
     on a staging/<name> branch in the infra repo
  4. issue_create — experiment brief (hypothesis + description + stack
     + provisioning log) on the new repo, returns the issue_url

Returns gitea_url, github_url, issue_url, next_steps. The next_steps
string is the exact shell sequence the operator runs locally to
clone, scaffold via local-dev 'task new-project', and push.

Idempotency: create_project_from_template + repo_mirror_push +
file_write_branch all return JSON-RPC code -32003 (Conflict) when
their target already exists; the orchestrator swallows the conflict
and continues. Re-running on an existing repo restates the brief in
a fresh issue.

Error handling: on any non-conflict downstream failure the response
returns {reached: ["<step>",...], failed_step: "<step>"} alongside
a JSON-RPC error. No rollback — partial state stays so the operator
can resume manually.

New env vars (all optional except GITEA_MCP_URL):
  GITEA_MCP_URL    enables the tool
  GITEA_MCP_TOKEN  bearer auth for gitea-mcp
  GITEA_OWNER      default mathias
  GITHUB_OWNER     default mathiasb
  INFRA_REPO       default infra
  GITHUB_PAT       repo scope, used as mirror remote_password; never logged

Without GITEA_MCP_URL set, the tool is not registered and the
routing pod starts normally (degrades open).

internal/mcpclient/: new minimal JSON-RPC tools/call client with
bearer auth, used by project_create. Unwraps MCP's
content[0].text envelope and surfaces typed errors via mcpclient.Error.

Tests: table-driven against an httptest fake gitea-mcp covering happy
path (4-step success + correct PATCH-style arg shapes), idempotent
repo-exists, mirror failure (partial-success response with reached=
[create_repo] + failed_step=mirror), infra-commit failure (reached up
to mirror + failed_step=infra_commit), and validation errors.

Closes #10
2026-05-18 11:44:39 +02:00
Mathias Bergqvist
c7e0192486 feat(auth): add Dex JWT middleware to supervisor, routing pod, and brain MCP
All checks were successful
CI / Lint / Test / Vet (push) Successful in 13s
CI / Mirror to GitHub (push) Successful in 3s
Closes #6 on gitea.d-ma.be/mathias/hyperguild.

Dex is deployed at auth.d-ma.be. All three MCP servers now accept JWTs
issued by Dex in addition to static bearer tokens, enabling claude.ai
OAuth 2.0 integration without abandoning backward-compat CLI auth.

Changes:
- internal/auth/: new Validator (JWKS auto-refresh via lestrrat-go/jwx/v2),
  ProtectedResourceHandler (RFC 9728 /.well-known/oauth-protected-resource)
- internal/mcp/Server: adds optional *auth.Validator; checkAuth tries JWT
  first, then static token fallback; both-nil = auth disabled (unchanged default)
- cmd/supervisor, cmd/routing: construct Validator from DEX_ISSUER_URL +
  MCP_AUDIENCE env vars; register protected-resource handler when set
- ingestion/internal/auth/: same Validator + handler (separate module)
- ingestion/internal/mcp/BearerAuth: same JWT-or-static chain
- ingestion/cmd/server: same wiring pattern

New env vars (all optional; absent = static-token-only, same as before):
  DEX_ISSUER_URL   — Dex issuer URL (e.g. https://auth.d-ma.be)
  MCP_AUDIENCE     — expected aud claim (e.g. brain, supervisor)
  MCP_RESOURCE_URL — resource identifier for RFC 9728 metadata response

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-11 20:10:05 +02:00
Mathias Bergqvist
5b207425ed refactor(routing): rename local/claude to fast/thinking model pair
All checks were successful
CI / Lint / Test / Vet (pull_request) Successful in 10s
CI / Mirror to GitHub (pull_request) Has been skipped
The routing decision is about reasoning capacity, not cost or provider.
Fast model (koala/qwen35-9b-fast) handles high-pass-rate calls; thinking
model (iguana/gemma4-26b) handles low-pass-rate calls. Removes the
implicit Anthropic dependency from the routing pod — both models go
through LiteLLM.

Renames: HYPERGUILD_LOCAL_MODEL → HYPERGUILD_FAST_MODEL,
HYPERGUILD_CLAUDE_MODEL → HYPERGUILD_THINKING_MODEL,
Router.LocalModel → FastModel, Router.ClaudeModel → ThinkingModel,
log decision "claude_fallback" → "thinking_fallback".

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-08 16:39:42 +02:00
Mathias Bergqvist
bee4bb3c1f chore(routing): pre-merge cleanup — Plan 7 reminders, code_review→review, operator note
All checks were successful
CI / Lint / Test / Vet (push) Successful in 11s
CI / Mirror to GitHub (push) Successful in 4s
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-05 23:22:15 +02:00
Mathias Bergqvist
083c2d7db9 feat(routing): cmd/routing binary
Wires Config → LiteLLMExecutor → Router → four skills (review, debug,
retrospective, trainer) → Registry → MCP server with bearer auth and
/healthz. Each skill's CompleteFunc is wrapped so the Router decides
local-vs-Claude per call and logs every decision to the brain /mcp.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-04 23:43:59 +02:00