looply-context-sync

name	looply-context-sync
description	Use to verify and sync context files (context-index, project-context, session-context, code-context) against the real codebase state.

Use this skill when the user explicitly invokes $looply-context-sync, asks to run /looply:context-sync, or clearly requests context synchronization.

Quick usage:

• $looply-context-sync [feature-name]

Primary references:

• Workflow playbook: ../../../../.looply/state/workflow-playbook.opencode.md
• Host status contract: ../../../../.looply/state/host-status-contract.json
• Managed pack: ../../../../.looply/managed/packs/software-delivery-suite
• Custom overrides: ../../../../.looply/custom
• Execution hints: ../../../../.looply/state/execution-hints.opencode.json
• Context index: ../../../../.looply/state/context-index.md
• Project context: ../../../../.looply/custom/project-context.md
• Session context: ../../../../.looply/custom/session-context.md
• Code context: ../../../../.looply/state/code-context.json
• Knowledge graph: ../../../../.looply/state/knowledge-graph.json

Usage:

• Explicit mention: $looply-context-sync
• Workflow alias to honor: /looply:context-sync
• Syntax in Codex: $looply-context-sync [feature-name]

Example:

• $looply-context-sync pix-webhook-retry

When to use:

• Context files are marked as stale, draft or empty
• Before starting a new workflow to ensure updated context
• After significant codebase changes (refactor, new dependency, architecture change)
• When auto-resume detects inconsistent context

Curated example guidance:

• ICL mode: on
• Use examples only for style, structure and quality calibration.
• Do not copy feature-specific names, identifiers or business details from examples.
• No example was selected for this workflow.

Context Sync Flow

Stage 1: Context Audit

1. Read all context files listed in Primary references.
2. Check the status of each one:
   • status: empty → file exists but has no content
   • status: draft → partial content, may be outdated
   • status: stale → content present but potentially outdated
   • status: valid → content is up to date
3. Compare modification timestamps with git log to detect staleness.

Stage 2: Codebase Validate

1. For each file with stale/draft/empty status, validate against the real codebase:
   • context-index.md → verify all listed features have a corresponding workflow-status.md
   • project-context.md → verify directory structure, dependencies in package.json, frameworks
   • session-context.md → verify session-links.json
   • code-context.json → run looply refresh-code-context if missing or stale
   • knowledge-graph.json → run looply refresh-code-context if missing or stale
2. Document discrepancies found.

Stage 3: Sync Report

Generate a report in .looply/custom/context-sync-report.md with:

# Context Sync Report

## Verified Files
| File | Status | Action |
|---------|--------|------|
| context-index.md | valid | keep |
| project-context.md | stale | update |

## Discrepancies Found
- project-context.md: dependency X not listed

## Suggested Actions
- Update project-context.md with dependency X
- Run looply refresh-code-context

Stage 4: Apply Sync

Present the report to the user and ask: "Context files need updating. Apply the fixes?" [y] Yes - apply all updates [n] No - keep as is [p] Partial - choose which to update

Execution rules:

1. Start by reading the workflow playbook and the context index file.
2. If the user asked for help, explain syntax, arguments, example, expected output and next step without mutating state.
3. Never overwrite manual context without user confirmation.
4. For code-context.json and knowledge-graph.json, use looply refresh-code-context.
5. Use managed pack files as canonical process definition and write local state only under .looply/custom.
6. Respond in the same language as the user's input.
7. For existing projects, use the real local codebase as the primary source of truth and use context files only as accelerators.
8. If a context file has status: empty, status: draft or status: stale, validate it against the local codebase before trusting it.
9. Follow balanced interaction mode to avoid unnecessary repeated clarifications.
10. When curated examples are referenced, use them only for style, structure and quality calibration.
11. Keep the response visually structured with clear Markdown section titles.
12. Do not use emojis.

---

Composed Agent Context

The sections below were pre-composed by looply from agent context_slots.

Constraints

• Never modify context files without user confirmation
• Always use looply refresh-code-context for code-context and knowledge-graph
• Do not modify managed files (only .looply/custom)

Escalation

• Escalate context inconsistencies to delivery-orchestrator
• Escalate codebase structure changes to architect Arguments:
• feature-name: optional short identifier of the feature to sync context for (optional)