WIP: feat/agent-desired-state #260

Draft

johnride wants to merge 2 commits from feat/agent-desired-state into master

Author	SHA1	Message	Date
Jean-Gabriel Gill-Couture	51b39505bb	docs(adr): reframe desired-state ADR as proposal and explore alternatives Some checks failed Run Check Script / check (pull_request) Failing after 37s Details Re-frame ADR-021 from an accepted shell-executor decision into an explicit problem statement plus one candidate proposal (Alternative A), with an Open Questions section capturing the concerns raised during review: wrong abstraction level, no idempotency, no resource model, no typed status, incoherence with the Score-Topology-Interpret pattern, and weak security posture. Add ADR-022 enumerating four alternatives: - A: shell command executor (current scaffold) - B: mini-kubelet with typed resource manifests and reconcilers - C: embedded Score interpreter on the agent - D: hybrid — typed manifests now, Scores later Recommends Alternative D: ship typed AgentManifest/AgentStatus with a small fixed reconciler set for the IoT MVP, keeping an explicit migration seam to the Score-based end state once Scores become uniformly wire-serializable. Also documents what specifically is wrong with the happy-path shell executor in harmony_agent/src/desired_state.rs and clarifies that the NATS KV watch + typed CAS write skeleton is reusable, while the execute_command shell-out should be gated behind an audited ShellJob variant or deleted once real reconcilers land.	2026-04-10 07:13:38 -04:00
Jean-Gabriel Gill-Couture	9cd1713788	feat(agent): desired-state convergence happy path Adds a proof-of-concept desired-state convergence mechanism for harmony_agent: the central platform writes a DesiredStateConfig to NATS KV at desired-state.<agent-id>; the agent watches the key, executes the command via sh -c, and writes the result to actual-state.<agent-id>. - New module: harmony_agent/src/desired_state.rs (~150 lines) - New types: DeploymentConfig::DesiredState, DesiredStateConfig, ActualState, ExecutionStatus - ADR: docs/adr/021-agent-desired-state-convergence.md This is an initial happy path for review. The architecture needs further evaluation against the mini-kubelet vision before hardening (security, sandboxing, signing, resource limits).	2026-04-10 07:08:53 -04:00

Author

SHA1

Message

Date

Jean-Gabriel Gill-Couture

51b39505bb

docs(adr): reframe desired-state ADR as proposal and explore alternatives

Run Check Script / check (pull_request) Failing after 37s

Details

Re-frame ADR-021 from an accepted shell-executor decision into an
explicit problem statement plus one candidate proposal (Alternative A),
with an Open Questions section capturing the concerns raised during
review: wrong abstraction level, no idempotency, no resource model, no
typed status, incoherence with the Score-Topology-Interpret pattern,
and weak security posture.

Add ADR-022 enumerating four alternatives:
- A: shell command executor (current scaffold)
- B: mini-kubelet with typed resource manifests and reconcilers
- C: embedded Score interpreter on the agent
- D: hybrid — typed manifests now, Scores later

Recommends Alternative D: ship typed AgentManifest/AgentStatus with a
small fixed reconciler set for the IoT MVP, keeping an explicit
migration seam to the Score-based end state once Scores become
uniformly wire-serializable.

Also documents what specifically is wrong with the happy-path shell
executor in harmony_agent/src/desired_state.rs and clarifies that the
NATS KV watch + typed CAS write skeleton is reusable, while the
execute_command shell-out should be gated behind an audited ShellJob
variant or deleted once real reconcilers land.

2026-04-10 07:13:38 -04:00

Jean-Gabriel Gill-Couture

9cd1713788

feat(agent): desired-state convergence happy path

Adds a proof-of-concept desired-state convergence mechanism for
harmony_agent: the central platform writes a DesiredStateConfig
to NATS KV at desired-state.<agent-id>; the agent watches the key,
executes the command via sh -c, and writes the result to
actual-state.<agent-id>.

- New module: harmony_agent/src/desired_state.rs (~150 lines)
- New types: DeploymentConfig::DesiredState, DesiredStateConfig,
  ActualState, ExecutionStatus
- ADR: docs/adr/021-agent-desired-state-convergence.md

This is an initial happy path for review. The architecture
needs further evaluation against the mini-kubelet vision before
hardening (security, sandboxing, signing, resource limits).

2026-04-10 07:08:53 -04:00

WIP: feat/agent-desired-state #260

2 Commits