data-ai MCP Server
Hook-based token compressor for 5 AI CLI hosts (Claude Code, Copilot CLI, OpenCode, Gemini CLI, Codex CLI). Up to 95% bash compression, signature-mode for code reads, cross-call dedup, MCP server, self-teaching protocol. Zero runtime deps.
Discovered via github-topic:mcp-server and last synced 3mo ago.
Install instructions not detected yet
Check the source repository for the latest setup steps.
Description
3,348 tk
50 tk
How
40,373 tk
3,402 tk
Baseline / month
524 tk
Last N bash invocations with hash + length + cmd snippet — check before re-running
Auto-teach payload — read once per session to learn squeez's markers + memory protocol
424 tk
640 tk
Token accounting + call counts (tokens_bash / tokens_read / tokens_other / seen_files / seen_errors / seen_git_refs)
Config path
Before
2,692 tk
Returns
Session efficiency scores: compression ratio, tool choice, context reuse, budget conservation (basis points)
82,257 tk
2,106 tk
502 tk
Files this session has touched, with access type (Read/Write/Created/Deleted), sorted by recency
1,881 tk
2,181 tk
441 tk
Error fingerprints with the first 128 chars of message text — find *what* the error was
`find`, `ls`, `du`, `ps`, `env`, `lsof`, `netstat`
`node`, `python`, `ruby`
2,306 tk
4,692 tk
182 tk
187 tk
Full structured view of a past session by date: calls, files, errors, git events, test summary
`curl`, `wget`
Memory file
731 tk
486 tk
Value
Sub-agent usage: spawn count, cumulative estimated tokens, per-call breakdown
Scaling
Latency
4,418 tk
4,957 tk
316 tk
717 tk
1,513 tk
full
665 tk
451 tk
Sessions where a given file path was touched, with token-savings and commit status
`psql`, `prisma`, `mysql`, `drizzle-kit`
Last N finalized prior-session summaries with structured fields: investigated / learned / completed / next_steps
Commands
`git`
1,782 tk
Distinct error fingerprints observed this session (FNV-1a hashes of normalized errors)
Key wins
Python
Dedup hit counts (exact + fuzzy), summarize triggers, Ultra-mode calls, tokens saved per category
Full-text search across all session summaries — find when you last saw an error or touched a file