AgentStack
SKILL verified MIT Self-run

Cvg Plan Loop

skill-gomilesf-convergo-cvg-plan-loop · by gomilesf

Orchestrate a monitored planning loop where a planner drafts the plan and fresh plan reviewers gate the exit. Use when the user asks to create or revise a plan through supervised specialist sessions on Codex threads or Claude Code background agents.

No reviews yet
0 installs
2 views
0.0% view→install

Install

$ agentstack add skill-gomilesf-convergo-cvg-plan-loop

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Cvg Plan Loop? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Plan Loop

Use this skill for plan-only cvg-multi-session orchestration.

Before doing anything, open and read references/cvg-multi-session-protocol.md, then apply the protocol gates from cvg-multi-session:

  • transport binding gate,
  • real specialist session gate,
  • orchestrator callback transport gate,
  • waiting handoff gate,
  • cvg-plan-review-feedback gate,
  • fresh-reviewer exit gate.

Do not implement product code in this workflow.

Phase 0: Orchestrator Setup

Record:

  • current orchestrator session id (Codex thread id; on Claude Code the harness tracks it),
  • worktree or repo path,
  • current git rev-parse HEAD if available,
  • dirty worktree state,
  • source prompt or input document,
  • expected plan path or target docs directory when known,
  • the loop ledger path, created per the Gate 5 loop-ledger rule.

Classify dirty files:

  • plan-owned files,
  • unrelated user artifacts,
  • files the planner must not touch.

Phase 1: Planner Handoff

Spawn the planner as a real specialist session (Gate 0 spawn operation) and record the id returned by the tool. Verify it per Gate 1; if it is not verifiable, do not continue.

(Codex) After verification, send the planner its verified thread id before creating the heartbeat.

Planner prompt must include:

  • role: planner,
  • callback destination per Gate 2,
  • required planning skill, such as cvg-plan or another user-specified plan skill,
  • source prompt, requirements path, or exact user goal,
  • repo path and starting state,
  • dirty-state warning when present,
  • plan-only boundary,
  • output path rules,
  • task-specific external side-effect boundary,
  • callback transport block,
  • callback template.

Do not restate the planning skill's process or paste broad repo rules, likely files, or local policy summaries unless they are the user's source authority.

Planner callback template:

I am the planner. My specialist id is . This planning round is complete. Plan path: . Key status: . Please decide the next step.

Complete the waiting handoff per Gate 3 and end the active turn.

Phase 2: Fresh Plan Review

After the planner callback is visible, spawn a new fresh reviewer specialist session and verify its id per Gate 1.

Reviewer prompt must include:

  • role: fresh plan reviewer,
  • the line Do not consult project memory, prior sessions, rollout summaries, or external history. The repository's own git history (log, blame, commit messages) is repo evidence, not external history. before the required skill line,
  • callback destination per Gate 2,
  • required cvg-plan-review skill, such as cvg-plan-review,
  • original planning goal or source prompt,
  • plan path,
  • repo path,
  • read-only boundary,
  • blocker-only reporting rule,
  • task-specific external side-effect boundary,
  • callback transport block,
  • callback template.

Do not paste the full planning prompt or all context documents when the plan links its authorities and the reviewer can read the worktree.

Reviewer callback template:

I am the fresh reviewer. My specialist id is . This first-pass full review is complete. Verdict: . Findings: . Please decide the next step.

Complete the waiting handoff per Gate 3 and end the active turn while waiting.

Blocking plan findings include:

  • missing required surface,
  • incomplete behavior contract coverage,
  • wrong ownership boundary,
  • unclear external or API contract,
  • missing migration, rollback, deletion, privacy, or safety lifecycle,
  • insufficient test or verification gate,
  • plan criterion that cannot be implemented as written,
  • unsafe side-effect assumption.

Phase 3: Return Review Feedback to Planner

If the reviewer reports blockers, send the reviewer feedback to the same verified planner session with the continue operation (Gate 0).

Do not classify findings, choose revision strategy, filter reviewer output, or turn the review into an edit list in the orchestrator.

The feedback prompt must include:

  • callback destination per Gate 2,
  • reviewer specialist id,
  • required feedback skill: cvg-plan-review-feedback,
  • original planning goal and source prompt,
  • current plan path and related contracts,
  • exact reviewer blocker findings appended under a Plan Review Feedback Input section,
  • task-specific external side-effect boundary,
  • callback transport block,
  • callback template.

Do not restate the feedback skill's intake, classification, or revision rules.

Complete the waiting handoff per Gate 3 and end the active turn.

Phase 4: Focused Re-Review

After the planner feedback callback is visible and a reviewable plan revision exists, send a focused re-review request to the same verified reviewer session with the continue operation.

Focused scope:

  • verify old blockers were addressed,
  • check directly related contradictions introduced by the revision,
  • do not redo broad review unless the plan's core design changed.

Complete the waiting handoff per Gate 3 and end the active turn.

If the planner reports a contract decision, escalation, or clarification need, route that callback to the appropriate reviewer or user decision before requesting re-review.

If same reviewer still finds blockers, repeat Phase 3.

If same reviewer passes, do not exit. Continue to Phase 5.

Phase 5: Final Fresh Review

Spawn a new fresh reviewer specialist session for a complete first review of the revised plan. Verify its id per Gate 1.

Use the same minimal reviewer prompt shape from Phase 2. The new fresh reviewer must deliver its callback to the orchestrator per Gate 2; do not accept a result left only inside the specialist session when the platform requires an explicit callback send.

Final exit condition:

  • new fresh reviewer,
  • complete first review,
  • no blocking findings.

Only then can the orchestrator declare the plan loop complete.

If the final fresh reviewer finds blockers, repeat from Phase 3. Each Phase 3-5 cycle is one round; after 3 rounds without a clean final review, stop and escalate to the user with all open findings and their adjudications (Gate 5 round cap). Apply the Gate 5 adjudication ratchet: findings previously adjudicated invalid or out of scope cannot re-block without new evidence.

Completion Summary

Report:

  • plan path,
  • planner specialist id,
  • reviewer specialist ids,
  • callback transport status,
  • blockers and planner cvg-plan-review-feedback results,
  • final fresh reviewer verdict,
  • heartbeat cleanup (Codex),
  • known gaps.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.