cleveragents/cleveragents-core

Fork 3

UAT: agents plan tree --format json output missing spec envelope and nested tree structure #5040

New issue

Closed

opened 2026-04-09 00:50:11 +00:00 by HAL9000 · 1 comment

HAL9000 commented

2026-04-09 00:50:11 +00:00

Owner

Bug Report

Feature Area: Plan Lifecycle — agents plan tree

Summary

agents plan tree <PLAN_ID> --format json returns a flat list of tree node dicts instead of the spec-required JSON envelope with a nested tree structure, summary, child_plans, and decision_ids sections.

Expected Behavior (from spec)

Per docs/specification.md §agents plan tree, the JSON output must be:

{
  "command": "plan tree",
  "status": "ok",
  "exit_code": 0,
  "data": {
    "plan_id": "01HXM8C2ZK4Q7C2B3F2R4VYV6J",
    "tree": {
      "type": "prompt_definition",
      "description": "Increase test coverage to 85%",
      "children": [
        { "type": "invariant_enforced", "description": "..." },
        { "type": "strategy_choice", "description": "...", "confidence": 0.82 },
        {
          "type": "subplan_parallel_spawn",
          "description": "...",
          "children": [
            { "type": "subplan_spawn", "description": "...", "plan_id": "01HXM9F1A" }
          ]
        }
      ]
    },
    "summary": {
      "nodes": 9,
      "depth": 3,
      "child_plans": "2+",
      "invariants": 2,
      "superseded": 0
    },
    "child_plans": [
      { "id": "01HXM9F1A", "phase": "execute", "state": "processing" }
    ],
    "decision_ids": {
      "root": "01HXM9A0B1Q2W3R5G8Z0P4Q1X8",
      "strategy": "01HXM9A1C2Q7W3R5G8Z0P4Q1X9"
    }
  },
  "timing": { "started": "2026-02-08T12:58:00Z", "duration_ms": 85 },
  "messages": ["Decision tree rendered"]
}

Key aspects of the spec's JSON tree structure:

data.tree is a nested structure (root node with children arrays)
Each node has type, description, and optionally confidence and plan_id
data.summary provides aggregate statistics
data.child_plans lists spawned child plans with their current status
data.decision_ids provides a named map of key decision IDs for correction

Actual Behavior

In src/cleveragents/cli/commands/plan.py, the tree_decisions_cmd command (lines 3994–4121) calls build_decision_tree() which returns a list of node dicts, then passes it directly to format_output. This produces:

No top-level envelope: No command, status, exit_code, timing, messages fields — just a bare JSON array
Different node structure: Nodes have decision_id, type, sequence, question, chosen, confidence, superseded, children — not the spec's type, description, confidence, plan_id
No data.summary section: Missing node count, depth, child_plans count, invariants count, superseded count
No data.child_plans section: Missing child plan status information
No data.decision_ids section: Missing the named map of key decision IDs for correction

The spec's tree node uses description (the chosen option text) while the implementation uses question and chosen as separate fields. The spec's type field matches, but the spec uses human-readable type names while the implementation uses the raw enum value.

Code Location

src/cleveragents/cli/commands/plan.py:

tree_decisions_cmd function (lines 3994–4121)
build_decision_tree function (lines 3918–3991)

Impact

Programmatic consumers of agents plan tree --format json receive a non-spec-compliant response. The missing decision_ids section is particularly impactful as it's the primary way users discover decision IDs for use with agents plan correct.

Automated by CleverAgents Bot
Supervisor: UAT Testing | Agent: uat-tester

## Bug Report **Feature Area:** Plan Lifecycle — `agents plan tree` ### Summary `agents plan tree <PLAN_ID> --format json` returns a flat list of tree node dicts instead of the spec-required JSON envelope with a nested tree structure, summary, child_plans, and decision_ids sections. ### Expected Behavior (from spec) Per `docs/specification.md` §agents plan tree, the JSON output must be: ```json { "command": "plan tree", "status": "ok", "exit_code": 0, "data": { "plan_id": "01HXM8C2ZK4Q7C2B3F2R4VYV6J", "tree": { "type": "prompt_definition", "description": "Increase test coverage to 85%", "children": [ { "type": "invariant_enforced", "description": "..." }, { "type": "strategy_choice", "description": "...", "confidence": 0.82 }, { "type": "subplan_parallel_spawn", "description": "...", "children": [ { "type": "subplan_spawn", "description": "...", "plan_id": "01HXM9F1A" } ] } ] }, "summary": { "nodes": 9, "depth": 3, "child_plans": "2+", "invariants": 2, "superseded": 0 }, "child_plans": [ { "id": "01HXM9F1A", "phase": "execute", "state": "processing" } ], "decision_ids": { "root": "01HXM9A0B1Q2W3R5G8Z0P4Q1X8", "strategy": "01HXM9A1C2Q7W3R5G8Z0P4Q1X9" } }, "timing": { "started": "2026-02-08T12:58:00Z", "duration_ms": 85 }, "messages": ["Decision tree rendered"] } ``` Key aspects of the spec's JSON tree structure: - `data.tree` is a **nested** structure (root node with `children` arrays) - Each node has `type`, `description`, and optionally `confidence` and `plan_id` - `data.summary` provides aggregate statistics - `data.child_plans` lists spawned child plans with their current status - `data.decision_ids` provides a named map of key decision IDs for correction ### Actual Behavior In `src/cleveragents/cli/commands/plan.py`, the `tree_decisions_cmd` command (lines 3994–4121) calls `build_decision_tree()` which returns a list of node dicts, then passes it directly to `format_output`. This produces: 1. **No top-level envelope**: No `command`, `status`, `exit_code`, `timing`, `messages` fields — just a bare JSON array 2. **Different node structure**: Nodes have `decision_id`, `type`, `sequence`, `question`, `chosen`, `confidence`, `superseded`, `children` — not the spec's `type`, `description`, `confidence`, `plan_id` 3. **No `data.summary` section**: Missing node count, depth, child_plans count, invariants count, superseded count 4. **No `data.child_plans` section**: Missing child plan status information 5. **No `data.decision_ids` section**: Missing the named map of key decision IDs for correction The spec's tree node uses `description` (the chosen option text) while the implementation uses `question` and `chosen` as separate fields. The spec's `type` field matches, but the spec uses human-readable type names while the implementation uses the raw enum value. ### Code Location `src/cleveragents/cli/commands/plan.py`: - `tree_decisions_cmd` function (lines 3994–4121) - `build_decision_tree` function (lines 3918–3991) ### Impact Programmatic consumers of `agents plan tree --format json` receive a non-spec-compliant response. The missing `decision_ids` section is particularly impactful as it's the primary way users discover decision IDs for use with `agents plan correct`. --- **Automated by CleverAgents Bot** Supervisor: UAT Testing | Agent: uat-tester

HAL9000 referenced this issue

2026-04-09 00:50:45 +00:00

[AUTO-UAT-POOL] UAT Testing Report (Cycle 2) #4987

HAL9000 added the

labels

2026-04-09 00:57:03 +00:00

HAL9000 commented

2026-04-09 00:58:06 +00:00

Author

Owner

Closing as duplicate of #4974 — both issues report the same problem: agents plan tree --format json output missing spec-required envelope.

Automated by CleverAgents Bot
Supervisor: Backlog Grooming | Agent: backlog-groomer

Closing as duplicate of #4974 — both issues report the same problem: `agents plan tree --format json` output missing spec-required envelope. --- **Automated by CleverAgents Bot** Supervisor: Backlog Grooming | Agent: backlog-groomer

HAL9000 closed this issue

2026-04-09 00:58:06 +00:00

HAL9000 referenced this issue

2026-04-09 00:58:39 +00:00

[AUTO-GROOMER] Backlog Grooming Report (Cycle 5) #5088

HAL9000 referenced this issue

2026-04-13 04:58:28 +00:00

Fix fail_fast cancellation in SubplanExecutionService #7807

HAL9000 referenced this issue

2026-04-13 08:38:41 +00:00

Fix fail_fast cancellation in SubplanExecutionService #7807

HAL9000 referenced this issue