goosedump

goosedump browses, searches, converts, and compacts saved coding-agent conversations. It supports Claude Code, Codex, Crush, Gemini CLI, Goose, OpenCode, and Pi.

It builds on ideas from lllyasviel/VCC.

Install

Prebuilt binaries are available for Linux x64 and arm64, macOS arm64, and Windows x64:

npm install --global @jarkkojs/goosedump

To build from source:

cargo build --release

Source builds require a C compiler for the bundled SQLite library.

Quick start

goosedump session list --query '.provider == "pi"'
goosedump session show pi:context-id
goosedump session find pi:context-id '*database*'
goosedump session search pi:context-id 'database failure'
goosedump session compact pi:context-id

Every command that operates on a saved session takes an explicit provider:session-id target. Use goosedump --help or the manual page for complete usage.

Commands

The command tree mirrors the things it manages:

session list [--query QUERY] [--json]
session show PROVIDER:ID [--entry ID] [--all] [--from ID] [--before ID] [--as PROVIDER]
session find PROVIDER:ID GLOB [range options] [--as PROVIDER]
session search PROVIDER:ID QUERY [range options] [--page NUMBER] [--as PROVIDER]
session compact PROVIDER:ID [range options] [--previous-summary SUMMARY] [--plain] [--remember] [--as PROVIDER]
session copy PROVIDER:ID --to PROVIDER [--yes]
session remove PROVIDER:ID [--yes]

memory search QUERY [--provider PROVIDER] [--path PATH] [--type TYPE] [--limit NUMBER]
memory status
memory forget HASH|PROVIDER:ID

model pull

session copy and session remove print their plan by default. Pass --yes to carry out the change. This makes destructive and cross-provider operations safe to inspect without a separate dry-run mode.

Persistent memory

session compact emits a summary without retaining the selected range. Add --remember only when the selected content may become durable memory.

Use memory search to retrieve learned decisions, facts, preferences, procedures, and episodes; memory status to inspect the store; and memory forget with a hash printed by memory search or a provider:session-id to remove it. Forgetting creates durable tombstones so the same content is not learned again accidentally.

Persistent formats are versioned and never migrated. An incompatible cache is rebuilt, while an incompatible memory database is rejected and must be reinitialized explicitly. A prior schema is never transformed in place.

Inference performance

The custom GPT-OSS-20B CPU engine uses physical CPU cores, capped by the CPUs available to the process, by default. Override the compute thread count:

GOOSEDUMP_INFERENCE_THREADS=6 \
goosedump session compact pi:context-id

GPT-OSS-20B text generation continues to use goosedump's custom CPU engine. Download its weights ahead of time with goosedump model pull.

The bench feature builds a standalone harness that shells out to goosedump session compact with GOOSEDUMP_PROFILE_COMPACT=1 and reports end-to-end and per-stage timings:

# Edit benches/bench.txt to point at a real context, then:
cargo bench --features bench

Each case emits JSON with per-run timing (including model loading), median and p95 latency, and per-stage medians and p95s. Keep the machine, context range, and model cache unchanged when comparing configurations.

Pi extension

The bundled pi-goosedump extension adds session-history search, entry expansion, durable-memory recall, CWD-scoped session management, memory forget/status maintenance, and turn-driven goosedump compaction to Pi:

pi install npm:pi-goosedump

Development

cargo test --all-targets
cargo clippy --all-targets -- -D warnings
cargo fmt --check

The published Linux x64 and arm64 binaries are static glibc PIEs.

Licensing

1. pi-goosedump: Apache 2.0
2. goosedump: LGPL 2.1+

Third Party Attribution

1. GPT-OSS-20B: Apache 2.0
   • Downloaded on first use at run-time. The model is not distributed together with goosedump.