Skip to content

Commit d07dac0

Browse files
committed
analysis: triage and analyse ryandonofrio3/osgrep
- Vendor tools/osgrep as submodule (9f2faf7, v0.5.16) - Add references/ryandonofrio3-osgrep.md (triage) - Add analysis/ANALYSIS-ryandonofrio3-osgrep.md (source-reviewed) Key finding: MCP server is a non-functional stub returning tools:[] Search pipeline (FTS + Granite 30M dense + ColBERT 17M reranking) confirmed Benchmark harness absent from repo; 10-query CSV directionally supports claims - Add benchmarks/sources/ryandonofrio3-osgrep-repro.md (not reproducible) - Update README, ANALYSIS, REVIEWED, PUNCHLIST, REFERENCE_INDEX
1 parent a8e7145 commit d07dac0

10 files changed

Lines changed: 434 additions & 5 deletions

File tree

.gitmodules

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -46,3 +46,6 @@
4646
[submodule "tools/glitterkill-sdl-mcp"]
4747
path = tools/glitterkill-sdl-mcp
4848
url = https://github.com/GlitterKill/sdl-mcp
49+
[submodule "tools/osgrep"]
50+
path = tools/osgrep
51+
url = https://github.com/Ryandonofrio3/osgrep

ANALYSIS.md

Lines changed: 6 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -26,6 +26,7 @@ Research synthesis across all analyzed tools and papers. Updated as individual `
2626
| [rtk](analysis/ANALYSIS-rtk.md) | Claude Code hook-based CLI proxy; two-track filter pipeline (69 Rust handlers + 58 TOML filters) | 60–90% on dev commands (as reported; chars/4 heuristic) | None — passthrough proxy | `scripts/benchmark.sh` runnable; live fixtures; 80% improvement CI gate | v0.35.0; Apache-2.0; TOML filter correctness enforced at compile time | No direct peer — transparent CLI-proxy hook architecture is unique. Use if goal is passive token reduction on Claude Code dev commands without changing agent behavior. Complementary to all MCP-layer tools. |
2727
| [socraticode](analysis/ANALYSIS-giancarloerra-socraticode.md) | Qdrant-backed hybrid search (dense + BM25 via Qdrant RRF); AST-aware chunking (18+ languages); polyglot dependency graph | 61.5% (as reported; bytes not tokens; single live session; no harness) | None | No harness; 3 narrative case studies | Docker required; Qdrant v1.15.2+ required; MIT | Overlaps codebase-memory-mcp and code-review-graph for code search. Prefer code-review-graph (no Docker/Qdrant dep, stronger benchmarks). Use socraticode only if Qdrant is already in the infrastructure stack and RRF hybrid search is needed. |
2828
| [sdl-mcp](analysis/ANALYSIS-glitterkill-sdl-mcp.md) | LadybugDB knowledge graph; Symbol Cards (~100 tokens/symbol, LLM summary, ETag re-fetch); Iris Gate Ladder 4-rung escalation; Delta Packs blast-radius; SCIP compiler-grade edges; 38 tool surfaces | 81% `tools/list` overhead (gateway mode, as reported); no end-to-end session figure | Implicit: Iris Gate Ladder prompts cheapest-first retrieval; `max-cards` on slices | No harness; no end-to-end benchmark; all figures author-run | LLM summary cost at index time undisclosed; LadybugDB opaque (no schema/SQL/Cypher); source-available license; 12 languages (Rust indexer); 125 stars | **Watch.** Iris Gate Ladder + ETag conditional re-fetch are architecturally novel — the most disciplined context-escalation model in this survey. But no end-to-end token savings figure exists, LadybugDB is opaque, and LLM index-time costs are undisclosed. Overlaps codebase-memory-mcp and code-review-graph for graph-based code intelligence; prefer those (MIT, transparent storage, broader languages) until SDL-MCP provides a reproducible end-to-end benchmark and documents summary costs. |
29+
| [osgrep](analysis/ANALYSIS-ryandonofrio3-osgrep.md) | npm CLI; LanceDB vector store; Granite 30M dense + mxbai ColBERT 17M (int8) late-interaction reranking; tree-sitter AST chunking; FTS + vector → RRF → two-stage ColBERT pipeline | ~20% cost reduction / ~30% speedup (as reported; 10-query, single-codebase CSV; no answer quality assessment) | None — result set size is the bound | 10-query CSV (opencode corpus, cost-only); internal MRR harness (`eval.ts`, 70+ cases, self-referential); not reproduced | MCP server is a non-functional stub as of commit 9f2faf7; last push 2026-01-17; 1,128 stars; Apache-2.0 | Overlaps socraticode and codebase-memory-mcp for semantic code search. Unique: richest hybrid retrieval pipeline in the survey (dense + FTS + two-stage ColBERT) in a zero-external-service npm package. MCP stub rules out MCP-framework integration currently. Prefer socraticode if Qdrant is already in the stack and RRF hybrid search is sufficient; prefer codebase-memory-mcp if graph queries are needed. Use osgrep if the call-stack trace + skeleton compression commands are the target use case or if a pure-npm zero-dep installation is required. |
2930

3031
---
3132

@@ -49,7 +50,8 @@ Research synthesis across all analyzed tools and papers. Updated as individual `
4950
10. **n2-arachne** — read for the fixed-percentage budget allocation model; chars/3.5 heuristic and non-commercial license are the two primary risks.
5051
11. **jdocmunch-mcp** — read for the O(1) byte-offset retrieval pattern; note the savings accounting flaw (counts all sections, not returned) and opt-out telemetry before adopting.
5152
12. **socraticode** — Qdrant-backed hybrid search (dense + BM25 via RRF); Docker required; the 61.5% figure is bytes not tokens from a single session — important methodological caveat shared with jdocmunch.
52-
13. **qmd** — most sophisticated retrieval pipeline in this survey (8 steps: BM25 probe → LLM query expansion → vec → RRF → rerank); relevant primarily when the agent's knowledge base is markdown, not code.
53-
14. **caveman** — output-style compression is a different category from all others; useful when output verbosity rather than input retrieval is the token budget bottleneck.
54-
15. **Understand-Anything** — read if developer comprehension (domain mapping, not token reduction) is the target; different value proposition from every other tool in this list.
55-
16. **git-semantic-bun** — borderline scope; read only if semantic retrieval from git commit history is specifically needed.
53+
13. **osgrep** — read alongside socraticode; richest hybrid retrieval pipeline in the survey (FTS + dense + two-stage ColBERT reranking) with zero external service dependencies; key caveat: MCP server is a non-functional stub as of last commit, and the 10-query benchmark covers cost only with no answer quality assessment.
54+
14. **qmd** — most sophisticated retrieval pipeline in this survey (8 steps: BM25 probe → LLM query expansion → vec → RRF → rerank); relevant primarily when the agent's knowledge base is markdown, not code.
55+
15. **caveman** — output-style compression is a different category from all others; useful when output verbosity rather than input retrieval is the token budget bottleneck.
56+
16. **Understand-Anything** — read if developer comprehension (domain mapping, not token reduction) is the target; different value proposition from every other tool in this list.
57+
17. **git-semantic-bun** — borderline scope; read only if semantic retrieval from git commit history is specifically needed.

PUNCHLIST.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -24,6 +24,7 @@ Legend: `[ ]` pending · `[~]` in progress · `[x]` done · `[-]` skipped (see R
2424
- [x] danjdewhurst/git-semantic-bun — CLI; semantic vector search over git history; no MCP; borderline scope
2525
- [x] safishamsi/graphify — Python CLI + MCP skill; multi-modal knowledge graph (code/PDFs/images/video); Tree-sitter AST + Leiden clustering; reports 71.5× token reduction; 3.7k stars
2626
- [x] sdl-mcp (GlitterKill/sdl-mcp) — TypeScript MCP server; Symbol Cards + Iris Gate Ladder escalation + SCIP integration; 38 tool surfaces; 81% `tools/list` reduction (gateway mode, as reported); source-available; 125 stars
27+
- [x] osgrep (Ryandonofrio3/osgrep) — npm CLI + Claude Code plugin; LanceDB vector store, tree-sitter AST chunking, Granite 30M dense + mxbai ColBERT 17M reranking, dual FTS+vector hybrid search; 20% cost / 30% speedup (as reported, 10-query CSV); MCP server is a stub; 1,128 stars; Apache-2.0
2728

2829
## Papers
2930

README.md

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -43,7 +43,7 @@ Context management intersects both: retrieval feeds the context window; a poison
4343

4444
## Tools under investigation
4545

46-
15 tools analyzed; all source-verified against vendored repos.
46+
16 tools analyzed; all source-verified against vendored repos.
4747

4848
| Tool | Approach | Analysis |
4949
|---|---|---|
@@ -62,6 +62,7 @@ Context management intersects both: retrieval feeds the context window; a poison
6262
| [caveman](https://github.com/JuliusBrussee/caveman) | Claude Code skill enforcing compressed output style | [analysis](analysis/ANALYSIS-juliusbrussee-caveman.md) |
6363
| [Understand-Anything](https://github.com/Lum1104/Understand-Anything) | Multi-agent pipeline → structural + domain knowledge graph dashboard | [analysis](analysis/ANALYSIS-lum1104-understand-anything.md) |
6464
| [git-semantic-bun](https://github.com/danjdewhurst/git-semantic-bun) | Local vector index over git commit history; hybrid BM25 + semantic ranking | [analysis](analysis/ANALYSIS-danjdewhurst-git-semantic-bun.md) |
65+
| [osgrep](https://github.com/Ryandonofrio3/osgrep) | npm CLI; LanceDB vector store, tree-sitter AST chunking, Granite 30M dense + mxbai ColBERT 17M reranking, two-stage hybrid search | [analysis](analysis/ANALYSIS-ryandonofrio3-osgrep.md) |
6566

6667
## Papers under investigation
6768

REVIEWED.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -10,6 +10,7 @@ See `AGENTS.md` for the full triage workflow.
1010

1111
| Date | Name | Type | Disposition | Notes |
1212
|---|---|---|---|---|
13+
| 2026-04-14 | ryandonofrio3-osgrep | tool | analysis | npm CLI + Claude Code plugin; LanceDB vector store, tree-sitter AST chunking, ColBERT reranking, dual Code/Docs channels, skeleton+trace commands; 20% token reduction / 30% speedup (as reported, benchmark methodology undocumented); 1,128 stars; Apache-2.0 |
1314
| 2026-04-14 | glitterkill-sdl-mcp | tool | analysis | TypeScript MCP server; Symbol Cards (~100 tokens/symbol), Iris Gate Ladder (4-rung context escalation), SCIP integration, 38 tool surfaces; 81% `tools/list` reduction in gateway mode (as reported); source-available; 125 stars |
1415
| 2026-04-13 | graphify | tool | analysis | Python CLI + MCP skill; multi-modal (code/PDFs/images/video); Tree-sitter AST + LLM semantic extraction; NetworkX + Leiden graph; 71.5× token reduction on curated 52-file corpus (as reported); 3.7k+ stars; MIT |
1516
| 2026-04-10 | colbymchenry-codegraph | tool | analysis | Node.js/WASM MCP server; single codegraph_explore tool; created 2026-01-18; 94% fewer tool calls (own eval runner); README copies CRG's diagrams and benchmark table — treat with caution; 412 stars; MIT |

0 commit comments

Comments
 (0)