Skip to main content

Maintenance CLI for the Claude Code / OpenClaw file-based memory system.

Project description

memory-doctor

Maintenance CLI for the Claude Code / OpenClaw file-based memory system. Four verbs:

memory-doctor status              # read-only summary
memory-doctor lint                # find dead [[wiki-links]]; exit 1 if any
memory-doctor ingest [--apply]    # promote pending handoffs into cards
memory-doctor compact [--apply]   # flatten multi-line MEMORY.md entries into topic files

ingest and compact default to dry-run; pass --apply to actually write.

Install

pipx install git+https://github.com/solomonneas/memory-doctor

Or from a local clone:

git clone https://github.com/solomonneas/memory-doctor && cd memory-doctor
pipx install .

Requires Python 3.10+. No runtime dependencies beyond stdlib.

Configuration

What Flag Env Default
Memory dir (cards + MEMORY.md) --memory-dir PATH MEMORY_DOCTOR_MEMORY_DIR ~/.claude/projects/<project-scope>/memory
Handoffs dir --handoffs-dir PATH MEMORY_DOCTOR_HANDOFFS_DIR ~/.openclaw/workspace/.claude/memory-handoffs
MEMORY.md threshold (lines) --max-lines N MEMORY_DOCTOR_MAX_LINES 180

<project-scope> is the dash-prefixed home-dir path Claude Code uses to scope per-project memory (e.g. -home-alice for user alice, -home-bob for user bob). The defaults are tuned for the OpenClaw layout. Override via flags or env for other setups.

What each verb does

status

Prints memory dir path, card count, MEMORY.md line+byte count, threshold status, dead-link count, handoffs dir path, pending + processed counts, oldest pending age. Exits 0. --json for a structured payload.

lint

Walks every .md in the memory dir, extracts [[wiki-link]] references, checks whether each target exists. Reports dead links grouped by source file with a closest-match suggestion (Levenshtein distance ≤ 3). Exits 0 if zero dead links, 1 if any (so you can gate a pre-commit hook on it).

ingest

Sweeps the handoffs dir for unprocessed *.md files matching the standard handoff template. For each one:

  • Recommended memory action: create-card writes a new card to the memory dir; skips on conflict (use --force to overwrite)
  • Recommended memory action: update-card appends the suggested content to an existing card; errors if the target is missing
  • Recommended memory action: no-card just moves the handoff to processed/

Successful handoffs are moved into <handoffs-dir>/processed/. Dry-run by default; --apply writes.

compact

Reads MEMORY.md, counts lines. If above the threshold, identifies multi-line entries (bullets whose detail spans more than one line) and proposes flattening them: keep the one-liner in the index, append the detail to the target topic file under a ## From index (YYYY-MM-DD) section. Dry-run by default; --apply writes (topic files first, MEMORY.md last). Refuses if a target topic file is missing (would orphan content). Warns if compaction alone won't bring MEMORY.md under threshold.

Examples

# Daily morning check:
memory-doctor status

# Drain the inbox:
memory-doctor ingest --apply

# Bring MEMORY.md back under threshold:
memory-doctor compact --apply

# Pre-push hook:
memory-doctor lint

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

memory_doctor-0.1.0.tar.gz (22.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

memory_doctor-0.1.0-py3-none-any.whl (15.1 kB view details)

Uploaded Python 3

File details

Details for the file memory_doctor-0.1.0.tar.gz.

File metadata

  • Download URL: memory_doctor-0.1.0.tar.gz
  • Upload date:
  • Size: 22.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.3

File hashes

Hashes for memory_doctor-0.1.0.tar.gz
Algorithm Hash digest
SHA256 a95207f1a3e26c70855ca420c03e1ec61f999b75a74c992486a06f20cc2f3b5f
MD5 e4576f1930b936530b57a077663a986e
BLAKE2b-256 ddddc5dc30f695dff7a7b89e3c4db4801bdf32a357484889250055e193f3af6e

See more details on using hashes here.

File details

Details for the file memory_doctor-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: memory_doctor-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 15.1 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.3

File hashes

Hashes for memory_doctor-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 d6ac1e3e45ecd9344190d22a034106272864342caf6e92b75add0427bcae739c
MD5 69452a55a9c1b8e677a7a68211b24ef6
BLAKE2b-256 9674b1f5e67fc608f03563fcc062255703c65913be93a1079d042555aaaeaaf3

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page