clean-codex-archives

Safely inspect and remove locally archived Codex conversation data on Windows, macOS, and Linux, including archived rollout JSONL files, archived session rows and related logs in Codex SQLite databases, and stale session_index.jsonl entries. Use when the user asks to clear, purge, delete, or clean archived Codex conversations or their local history metadata.

Jump to install

Repository: LunarCache/skills
Last source activity: July 23, 2026 at 08:35
Detected SKILL.md language: English
Stars: 0
Forks: 0

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

A direct command skips the review prompt. Inspect the source before running it.

npx skills add https://github.com/LunarCache/skills --skill clean-codex-archives

The command stays on one line. Scroll horizontally to inspect it before copying.

Prefer a local copy? Download the files currently available to SkillsMP.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

File Explorer

3 files

SKILL.md

Source instructions · Read-only preview

Back to search Back to installation

skill-for-ragflow

Operate RAGFlow v0.26.4 deployments through the bundled Node CLI and API client. Use when user needs to manage RAGFlow datasets, documents, uploads, parsing, chunks, retrieval, chat assistants, chat sessions, agents, agent sessions, embedded website access, metadata filters, model discovery, model providers, system settings, or API diagnostics. Also use when the user asks about knowledge bases, document chunking, vector retrieval, embed code, or RAG workflows and the current context explicitly involves a RAGFlow server or deployment.

2026-07-23

View all from this repository

Clean Codex Archives

Use the bundled scripts/clean_codex_archives.py script as the cross-platform entry point. It uses only Python's standard library and preserves active sessions.

Safety contract

Resolve the Codex data root from an explicit --codex-home argument, then the CODEX_HOME environment variable, then the current user's .codex directory.

Treat state_5.sqlite as the source of truth. Only rows with threads.archived = 1 are eligible for deletion.

Preserve all unarchived threads, their logs, and their session-index entries.

Run without --apply first. Use --apply only after the user explicitly authorizes deletion.

Delete associated logs and goals before deleting archived thread rows, because the archived IDs are needed to select related records safely.

Delete related thread_dynamic_tools and thread_spawn_edges rows when those tables are present.

Remove only archived_sessions/*.jsonl files. Do not delete the archive directory itself or unrelated files.

Rewrite session_index.jsonl only after database cleanup, retaining entries whose IDs still exist in state_5.sqlite and rejecting malformed JSON instead of silently dropping it.

If a database is locked, ask the user to close Codex or otherwise stop the process holding the database, then retry.

Workflow

Run a dry run and inspect the counts for archive files, archived threads, related logs, goals, and stale index entries.

If the user confirms deletion, rerun with --apply.

Use --vacuum only when the user also wants SQLite files compacted; it may require exclusive database access and is not needed for logical deletion.

Verify that archived-thread count is zero, archive-file count is zero, remaining index IDs all exist in the state database, and SQLite quick_check reports ok for modified databases.

Commands

# Inventory only; no changes. Use python3 on Unix-like systems. python3 ./scripts/clean_codex_archives.py # Apply the explicitly authorized cleanup. python3 ./scripts/clean_codex_archives.py --apply # Use a specific Codex data root. python3 ./scripts/clean_codex_archives.py --codex-home "$HOME/.codex" --apply # Optional physical compaction after deletion. python3 ./scripts/clean_codex_archives.py --apply --vacuum

On Windows, use python instead of python3.