Plugin directory / AI Models / @kolawong/fast-compaction-dsh
@kolawong/fast-compaction-dsh
Unverified
4Stars
AI ModelsCategory
GitHubSource
What it does
Verdict-based context compaction for DeepSeek Harness — replaces lossy LLM summaries with fast keep/truncate/drop decisions from jev-latest; everything kept stays verbatim. Port of tamaratran/fast-jev-compaction.
Our take
Unverified — not yet install-tested; try it in your own environment.
Unverified — not yet install-tested; try it in your own environment.
“Verified” means our CI ran dsh plugin add in a clean profile and it booted — not a security audit and not an endorsement of third-party code.
Install
🧩 Let your agent install it (recommended)
Install the catalog once, then ask DeepSeek Harness to find and install this plugin.
dsh plugin add dshbase-catalogWeb profile:
dsh plugin --profile web add github:kolawong/fast-compaction-dsh