agentleFS
Sign inStart free
AlexsJonesAlexsJones/sympozium/CLAUDE.md

sympozium

CLAUDE.md675 starsChanged 12 days agoMITView on GitHub

Guidance for Claude Code working in this repo. The canonical contributor guide is AGENTS.md — read it for repo layout, CRD list, and common tasks. This file covers only what's easy to get wrong.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 1

    1. sympozium-ai/sympozium★ 675100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repositoryOne other file it gives its agents.

The file

We split this CLAUDE.md into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Always run make manifests (or make generate for deepcopy too).

    line 21
  2. Very specific:

    Commit or push only when asked.

    line 59

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

CLAUDE.md

Guidance for Claude Code working in this repo. The canonical contributor guide is AGENTS.md — read it for repo layout, CRD list, and common tasks. This file covers only what's easy to get wrong.

Build / test / verify

  • Go 1.25+. Module: github.com/sympozium-ai/sympozium.
  • make build — all binaries. go build ./... for a quick compile check.
  • make test runs go test -race ./... — the race detector is the bar; a change isn't done until go test -race ./..., go vet ./..., and gofmt are clean.
  • make test-system (envtest) runs controller tests against a real apiserver with no cluster — use it for reconcile-logic changes. CI runs it (system-tests job).
  • go vet ./... skips build-tagged files (test/system is //go:build system); make vet / make vet-tags covers them — add a line there for any new build tag.
  • Integration tests (test/integration/*.sh) need a Kind cluster + a model. They work against any OpenAI-compatible provider — set spec.model.provider + spec.model.baseURL (e.g. a local llama-server/ollama/lm-studio), not just openai.

After editing CRD types (api/v1alpha1/*)

Always run make manifests (or make generate for deepcopy too). This regenerates config/crd/bases/ and syncs both chart copies (charts/sympozium/crds/, charts/sympozium-crds/templates/). CI fails if they drift (make helm-sync-check). Never hand-edit generated CRD YAML or zz_generated.deepcopy.go.

Updating the Pi/Hermes example adapter versions

charts/sympozium/values.yaml's harnessExamples block is the only file to edit for an adapter's name, image digest, or default enablement — the --reuse-values fallback dict in charts/sympozium/templates/harness-examples.yaml is generated from it. Run make helm-sync after editing it; make helm-sync-check (CI) fails if the two have drifted. Never hand-edit the block between the sync-harness-defaults:begin/:end markers.

Naming — these renames recur as bugs

  • The CRD kind is Agent (formerly SympoziumInstance — do not use).
  • AgentRun/Schedule reference an agent via agentRef + agentId (not instanceRef/agentID).
  • Ensemble members live under agentConfigs (not personas).
  • The run-pod label is sympozium.ai/agent-run (hyphen, not agentrun).
  • AgentRun.spec.model requires provider, model, authSecretRef; toolPolicy actions are only allow/deny; cleanup is only delete/keep.

Conventions to match when adding code

  • Secrets → env: inject only via the allowedAuthSecretKeys allowlist with per-key SecretKeyRef. Never EnvFrom a whole secret into an agent/sidecar pod.
  • Pod security: agent pods run non-root, readOnlyRootFilesystem, drop: [ALL], RuntimeDefault seccomp. Match this for any new pod spec.
  • Admission: the webhook (cmd/webhook) is a separate, validation-only deployment — there is no wired mutating webhook, so don't assume field defaulting happens at admission; do it in the controller.
  • Image tags: control-plane deployments use a v-prefixed tag; spawned run pods use SYMPOZIUM_IMAGE_TAG unprefixed (_helpers.tpl vs the controller env).
  • Agents are LLM-driven and treated as adversarial (prompt injection) — validate anything an agent writes into /ipc or a CR field before acting on it.

Deploy / commit

  • helm lint charts/sympozium after chart changes; the two-chart split (sympozium-crds + sympozium) means CRDs upgrade independently.
  • Commit or push only when asked; branch off main first. Don't hand-commit generated artifacts — run make generate/manifests and commit the result.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free