patxkil/setkontext

View on GitHub by @patxkil

Engineering context for AI coding agents — decisions from GitHub + session memory (bugs, gotchas, implementations) via MCP

What's novel

Engineering context for AI coding agents — decisions from GitHub + session memory (bugs, gotchas, implementations) via MCP

Code Analysis

11 files read · 4 rounds

Extracts engineering decisions from GitHub repos (ADRs, PRs, docs) and AI coding session transcripts, stores them in SQLite with FTS5, and exposes them to AI coding agents via MCP tools for querying, validation, and session memory recall.

Strengths

The consolidation loop (promoting recurring session learnings into formal decisions) is a genuinely novel feedback mechanism for AI agent memory. The MCP server design is production-quality with tiered detail levels, graceful degradation without API keys, and activity logging.

Weaknesses

Entity extraction in ADRs relies on a hardcoded keyword list, and the FTS query builder is a simple stop-word + OR approach that may miss nuanced queries. LLM calls are synchronous with time.sleep for backoff, which blocks the MCP server under concurrent load.

Score Breakdown

Innovation

7 (25%)

Craft

76 (35%)

Traction

14 (15%)

Scope

75 (25%)

Signal breakdown

Innovation

Not Fork+1

Code Novelty+2

Concept Novelty+2

Craft

Ci+5

Tests+8

Polish+0

Releases+0

Has License+5

Code Quality+26

Readme Quality+15

Recent Activity+7

Structure Quality+5

Commit Consistency+0

Has Dependency Mgmt+5

Traction

Forks+0

Stars+6

Hn Points+0

Watchers+3

Early Traction+5

Devto Reactions+0

Community Contribs+0

Scope

Commits+7

Languages+3

Subsystems+10

Bloat Penalty+0

Completeness+7

Contributors+5

Authored Files+12

Readme Code Match+3

Architecture Depth+5

Implementation Depth+8

Evidence

Commits

Contributors

Files

Active weeks

TestsCI/CDREADMELicenseContributing

Repository

Language

Python

Stars

Forks

License

MIT