run_87396509fb13bbf1

input

{
  "request": "Triage Jira ticket MZCLD-3647: Add support for container workingDir in MozCloud Helm charts (jobs/workloads)",
  "references": [
    "https://mozilla-hub.atlassian.net/browse/MZCLD-3647"
  ],
  "sourceContext": {
    "issue": "MZCLD-3647"
  }
}

summary

MZCLD-3647 requests adding support for setting `containers[].workingDir` on workloads, Jobs, and CronJobs in the MozCloud Helm charts. This is a small, well-scoped feature request surfaced from the llm-proxy migration (MZCLD-2331), where the `litellm-migrations` and `mlpa-migrations` Jobs need `workingDir: /app` but the current chart schema has no way to express this field. The ticket is currently unassigned/untriaged in Backlog with no comments. Given clear acceptance criteria (workingDir configurable per container, appears in rendered pod spec) and a concrete unblocked-by-workaround dependency (MZCLD-2331 migration), this looks like a good candidate for prioritization into an upcoming sprint, likely a straightforward chart schema + template addition analogous to existing per-container fields (e.g., command/args/env) already supported.

findings

• Ticket is well-defined with clear problem statement, ask, and acceptance criteria; low ambiguity for engineering to pick up.
• This is a real, active blocker (or workaround-requiring gap) for the in-progress llm-proxy mozcloud migration, not a speculative feature request.
• Linked ticket MZCLD-2331 could not be read (blocked by policy: resource not in run subject), so full migration context/urgency/timeline could not be independently verified in this triage.
• No comments or prior triage activity exist on the ticket, and status is Backlog, suggesting it has not yet been scoped or assigned to an engineer.
• Scope spans multiple chart workload types (Deployments/workloads, Jobs, CronJobs), implying changes to multiple templates/values schemas plus documentation and tests, though the change itself (adding one optional per-container field) is small in complexity.

proposed actions

Triage notes: This request is well-scoped and tied to an active migration blocker (llm-proxy / MZCLD-2331 Jobs `litellm-migrations` and `mlpa-migrations` need `workingDir: /app`). Recommend prioritizing into the next sprint given it unblocks an in-flight migration. Suggested implementation approach: add an optional `workingDir` field to the per-container values schema (mirroring how `command`/`args`/`env` are already exposed) and thread it through the pod spec templates for Deployments/workloads, Jobs, and CronJobs. Please confirm scope covers all three workload types and add a values-schema validation + a chart unit/snapshot test asserting `workingDir` renders correctly in the pod spec. Also confirm with the llm-proxy migration owner (MZCLD-2331) whether this is currently blocking their rollout so we can set urgency accordingly.

tool calls

[
  {
    "capability": "jira.issue.read",
    "args": {
      "key": "MZCLD-3647"
    },
    "result": {
      "key": "MZCLD-3647",
      "summary": "Add support for container workingDir in MozCloud Helm charts (jobs/workloads)",
      "status": "Backlog",
      "description_text": "Context:  Surfaced during the  llm-proxy  mozcloud migration (MZCLD-2331). Problem.  Containers cannot set  workingDir . The  llm-proxy  migration Jobs ( litellm-migrations ,  mlpa-migrations ) both set  workingDir: /app ; this can't currently be expressed in the mozcloud chart. Ask.  Allow setting  containers[].workingDir  on workloads, jobs, and cronjobs. Acceptance criteria.   workingDir  is configurable per container and appears in the rendered pod spec.",
      "comment_count": 0,
      "comments": []
    }
  },
  {
    "capability": "jira.issue.read",
    "args": {
      "key": "MZCLD-2331"
    },
    "result": {
      "error": "read blocked by policy: resource not in run subject",
      "args": {
        "key": "MZCLD-2331"
      }
    },
    "blocked": true
  }
]

log