[Feature]:Saga-tree visualization: 'cleo saga tree <id>' showing recursive saga→epic→task hierarchy with phase/severity/depends inline

[kryptobaseddev]

Feature Request

Summary

cleo saga members <id> returns flat list of member epic IDs. Building the full saga→epic→task hierarchy (with each epic's phase, dependency edges, and child tasks) requires 7+ separate commands the operator must compose mentally. A cleo saga tree <id> command would make multi-saga orchestration tractable.

Concrete need

To validate the SG-SoR (T119) dep graph during the 2026-05-21 session, I had to run:

cleo show T119                                 # see saga relates[type=groups] members
cleo list --parent T123                        # E-SoR.1 children (14 ADRs)
cleo list --parent T124                        # E-SoR.6 children
cleo list --parent T127                        # E-SoR.5 children
cleo show T127 | jq '.data.task.depends'       # T127 dep chain
cleo show T128 | jq '.data.task.depends'       # T128 dep chain
cleo show T125 | jq '.data.task.depends'       # T125 dep chain
cleo show T126 | jq '.data.task.depends'       # T126 dep chain
cleo show T124 | jq '.data.task.depends'       # T124 dep chain

…then mentally reconstruct the tree. The manual phase-status table in NEXT_SESSION.md is the artifact of this exercise. Without a built-in tree view, each orchestrator session repeats it.

Proposed design

cleo saga tree T119
# Output:
T119 SG-SoR — Kodo as relational System of Record [saga, pending]
├── T123 E-SoR.1 — Research + ADRs           [phase:R+C, pending, 14 children]
│   ├── T129 ADR-SoR-006 Evidence-span schema
│   ├── T130 ADR-SoR-007 Supersedence edges
│   ├── ...
│   └── T148 ADR-SoR-011 Gemini migration   [priority:critical, deadline:2026-06-01]
├── T122 E-SoR.2 — Entity graph             [phase:A+S, pending, 5 children]
├── T121 E-SoR.3 — DSL + extraction          [phase:A+S, pending, 6 children]
├── T120 E-SoR.4 — Cache + retrieval         [phase:A+S, pending, 0 children]
├── T127 E-SoR.5 — Foundation impl          [phase:D, depends:T120,T121,T122,T123]
├── T124 E-SoR.6 — Extraction impl           [phase:D, depends:T123,T121]
│   └── T149 Gemini cutover                  [priority:critical, deadline:2026-06-01]
├── T125 E-SoR.7 — Moat capabilities         [phase:D, depends:T127]
├── T128 E-SoR.8 — Cache impl                [phase:D, depends:T160]   ← cross-saga edge to T009/E-F.6
└── T126 E-SoR.9 — Validation                [phase:V, depends:T127,T124,T125,T128]

Cross-saga edges:
  T128 depends T160 (from T009/SG-Foundation/E-F.6 redis service)

Flags

• --depth N — limit recursion
• --show depends,phase,priority,severity,deadline — opt-in columns
• --format text|json|mermaid — text-tree (default), JSON for programmatic, Mermaid for diagram embedding
• --cross-track — highlight edges that cross saga boundaries

Impact

The user's manual analysis required ~7 separate commands to reconstruct what cleo saga tree T119 would emit in one call. For an orchestrator dispatching N agents, this is a high-frequency operation.

Pairs naturally with the saga-traversal bug fix (#390) and the deps-show rendering improvement (#399) — both depend on the same recursive-traversal engine.

Cross-ref

Surfaced during SG-SoR validation 2026-05-21. Related: #390 (orchestrate ready/waves saga traversal), #399 (deps show rendering).
Area: cli

Are you using an AI agent?

Yes - AI agent filed this issue

---

Environment

Component	Version
CLEO	2026.2.1
Node.js	v24.13.1
OS	linux 6.19.14-200.fc43.x86_64 x64 (x64)
Shell	/bin/bash
gh CLI	gh version 2.87.3 (2026-02-23)
Install	/home/keatonhoskins/.npm-global/bin/cleo