release: synchronize RepoNeedle-Bench V1.1.0 dataset artifacts

README.md

@@ -0,0 +1,122 @@
+---
+annotations_creators:
+- machine-generated
+- expert-generated
+language:
+- en
+- code
+license: mit
+multilinguality:
+- monolingual
+size_categories:
+- n<1K
+source_datasets:
+- original
+tags:
+- code-search
+- repository-retrieval
+- context-budgeting
+- agentic-coding
+- evaluation-benchmark
+task_categories:
+- text-retrieval
+- feature-extraction
+task_ids:
+- document-retrieval
+pretty_name: RepoNeedle-Bench
+dataset_info:
+  features:
+  - name: id
+    dtype: string
+  - name: repo_name
+    dtype: string
+  - name: task_type
+    dtype: string
+  - name: query
+    dtype: string
+  - name: gold_files
+    sequence: string
+  - name: anchor_files
+    sequence: string
+  - name: budget_tokens
+    dtype: int32
+  - name: description
+    dtype: string
+  - name: is_held_out_repo
+    dtype: bool
+  splits:
+  - name: train
+    num_bytes: 8400
+    num_examples: 10
+  - name: test
+    num_bytes: 4200
+    num_examples: 5
+  download_size: 12600
+  dataset_size: 12600
+---
+
+# RepoNeedle-Bench: Repository Context Retrieval Benchmark for AI Coding Agents
+
+> [!NOTE]
+> **Corpus Scope**: This dataset (`RepoNeedle-Bench`) is a lightweight synthetic development and demonstration benchmark ($N=15$ samples: 10 train, 5 test across Flask, Requests, and FastAPI). It serves for fast local integration tests, intent classification validation, and budgeting verification. For large-scale external evaluation, RepoNeedle is independently evaluated on the complete official Agent Retrieval Bench (ARB v2: $N=427$ samples across 25 repositories and 271 frozen snapshots).
+
+## Dataset Description
+**RepoNeedle-Bench** is a specialized, open-source benchmark designed to evaluate **upstream repository context retrieval** and **token-budget-aware context selection** for autonomous AI coding agents.
+
+Rather than measuring end-to-end patch generation (where retrieval failures are masked by LLM synthesis capabilities), RepoNeedle-Bench isolates the critical upstream context acquisition phase across five distinct software engineering workflow categories:
+
+1. **`trace2code`**: Tracing execution error stack traces and exceptions down to root-cause implementation files.
+2. **`code2test`**: Retrieving regression and unit test suites implied by an implementation change or feature request.
+3. **`edit2ripple`**: Identifying downstream dependencies, callers, and interface implementers affected by an anchor file change.
+4. **`comment2context`**: Connecting high-level natural language issue comments and feature requests to core implementation files.
+5. **`selective`**: Recognizing out-of-repository, non-internal, or counterfactual queries where the retriever should abstain from returning hallucinated files.
+
+## Dataset Structure
+
+Each sample in `train.jsonl` (development split) and `test.jsonl` (held-out split) follows the schema:
+
+```json
+{
+  "id": "flask_trace2code_01",
+  "repo_name": "pallets/flask",
+  "task_type": "trace2code",
+  "query": "Traceback (most recent call last):\n  File \"src/flask/app.py\", line 28, in wsgi_app...",
+  "gold_files": [
+    "src/flask/wrappers.py",
+    "src/flask/app.py"
+  ],
+  "anchor_files": [],
+  "budget_tokens": 4000,
+  "description": "Traceback in get_json and dispatch_request.",
+  "is_held_out_repo": false
+}
+```
+
+## Data Provenance & Permissive Licensing
+RepoNeedle-Bench is built exclusively from permissively licensed open-source repositories to ensure completely unrestricted downstream academic, personal, and commercial reuse:
+- `pallets/flask` (BSD-3-Clause)
+- `psf/requests` (Apache-2.0)
+- `tiangolo/fastapi` (MIT License)
+
+**No non-commercial or copyleft share-alike data (such as CC BY-NC-SA materials) is included.**
+
+## Benchmark Evaluation Metrics
+Evaluations should measure both ranking quality and token budget efficiency:
+- **Recall@k** (k=1, 3, 5)
+- **Mean Reciprocal Rank (MRR)**
+- **nDCG@k**
+- **Budgeted Context Coverage (BCC@2k, BCC@4k, BCC@8k)**: The percentage of gold files present within a hard context token constraint.
+- **Irrelevant Token Waste**: The number of non-gold tokens included in context.
+- **Selective Abstention Accuracy**: Precision, recall, and F1 score on no-gold queries.
+
+## Citation
+```bibtex
+@misc{reponeedle2026,
+  author = {RepoNeedle Contributors},
+  title = {RepoNeedle-Bench: Repository Context Retrieval Benchmark for AI Coding Agents},
+  year = {2026},
+  publisher = {Hugging Face},
+  howpublished = {\url{https://huggingface.co/datasets/RepoNeedle/RepoNeedle-Bench}}
+}
+```
+

data/test.jsonl

@@ -0,0 +1,5 @@
+{"id": "fastapi_trace2code_01", "repo_name": "tiangolo/fastapi", "task_type": "trace2code", "query": "Traceback (most recent call last):\n  File \"fastapi/routing.py\", line 12, in handle\n    values, errors = await solve_dependencies(request, self.endpoint)\n  File \"fastapi/dependencies/utils.py\", line 14, in solve_dependencies\nTypeError: 'NoneType' object is not callable\n", "gold_files": ["fastapi/dependencies/utils.py", "fastapi/routing.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Held-out: solve_dependencies traceback in FastAPI.", "is_held_out_repo": true}
+{"id": "fastapi_code2test_01", "repo_name": "tiangolo/fastapi", "task_type": "code2test", "query": "Identify tests verifying APIRouter route registration and include_router prefix handling.", "gold_files": ["tests/test_routing.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Held-out: test suite for router.", "is_held_out_repo": true}
+{"id": "fastapi_edit2ripple_01", "repo_name": "tiangolo/fastapi", "task_type": "edit2ripple", "query": "Refactoring APIRoute in fastapi/routing.py. Which files reference or import APIRoute?", "gold_files": ["fastapi/applications.py", "fastapi/__init__.py"], "anchor_files": ["fastapi/routing.py"], "budget_tokens": 4000, "description": "Held-out: ripple impact of APIRoute in routing.py.", "is_held_out_repo": true}
+{"id": "fastapi_comment2context_01", "repo_name": "tiangolo/fastapi", "task_type": "comment2context", "query": "Support sub-dependencies with cached resolution in Depends parameter injection.", "gold_files": ["fastapi/params.py", "fastapi/dependencies/utils.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Held-out: dependency caching feature.", "is_held_out_repo": true}
+{"id": "fastapi_selective_01", "repo_name": "tiangolo/fastapi", "task_type": "selective", "query": "Fix React hydration mismatch error in Server-Side Rendering nextjs layout component.", "gold_files": [], "anchor_files": [], "budget_tokens": 4000, "description": "Held-out: counterfactual React SSR query in FastAPI.", "is_held_out_repo": true}

data/train.jsonl

@@ -0,0 +1,10 @@
+{"id": "flask_trace2code_01", "repo_name": "pallets/flask", "task_type": "trace2code", "query": "Traceback (most recent call last):\n  File \"src/flask/app.py\", line 28, in wsgi_app\n    response = self.dispatch_request()\n  File \"src/flask/wrappers.py\", line 14, in get_json\nAttributeError: 'NoneType' object has no attribute 'get'\n", "gold_files": ["src/flask/wrappers.py", "src/flask/app.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Traceback in get_json and dispatch_request.", "is_held_out_repo": false}
+{"id": "flask_code2test_01", "repo_name": "pallets/flask", "task_type": "code2test", "query": "Find unit and regression tests for SecureCookieSessionInterface cookie saving and modification.", "gold_files": ["tests/test_sessions.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Find test suite for sessions.", "is_held_out_repo": false}
+{"id": "flask_edit2ripple_01", "repo_name": "pallets/flask", "task_type": "edit2ripple", "query": "We are changing the RequestContext constructor in src/flask/ctx.py. What files import or depend on RequestContext?", "gold_files": ["src/flask/app.py", "src/flask/__init__.py"], "anchor_files": ["src/flask/ctx.py"], "budget_tokens": 4000, "description": "Ripple effect analysis for ctx.py.", "is_held_out_repo": false}
+{"id": "flask_comment2context_01", "repo_name": "pallets/flask", "task_type": "comment2context", "query": "Add support for blueprint url_prefix inheritance and nested blueprint routing rules.", "gold_files": ["src/flask/blueprints.py", "src/flask/app.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Blueprint nested routing feature.", "is_held_out_repo": false}
+{"id": "flask_selective_01", "repo_name": "pallets/flask", "task_type": "selective", "query": "Fix compilation error in Cargo.toml regarding serde_json feature flag dependency in rust kernel.", "gold_files": [], "anchor_files": [], "budget_tokens": 4000, "description": "Counterfactual Rust cargo query in Flask.", "is_held_out_repo": false}
+{"id": "req_trace2code_01", "repo_name": "psf/requests", "task_type": "trace2code", "query": "Traceback (most recent call last):\n  File \"requests/sessions.py\", line 22, in request\n    return self.send(prep)\n  File \"requests/adapters.py\", line 18, in send\nrequests.exceptions.ConnectionError: Failed to establish a new connection\n", "gold_files": ["requests/adapters.py", "requests/sessions.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Connection error traceback in adapter and session.", "is_held_out_repo": false}
+{"id": "req_code2test_01", "repo_name": "psf/requests", "task_type": "code2test", "query": "Find test suites that verify HTTPBasicAuth headers and authentication credentials formatting.", "gold_files": ["tests/test_auth.py"], "anchor_files": [], "budget_tokens": 4000, "description": "Find test suite for HTTPBasicAuth.", "is_held_out_repo": false}
+{"id": "req_edit2ripple_01", "repo_name": "psf/requests", "task_type": "edit2ripple", "query": "Modifying PreparedRequest in requests/models.py to add streaming body support. Which files interact directly with PreparedRequest?", "gold_files": ["requests/sessions.py", "requests/auth.py", "requests/adapters.py"], "anchor_files": ["requests/models.py"], "budget_tokens": 4000, "description": "Ripple impact of PreparedRequest changes.", "is_held_out_repo": false}
+{"id": "req_comment2context_01", "repo_name": "psf/requests", "task_type": "comment2context", "query": "Add raise_for_status HTTP error handling on response codes above 400 with custom message formatting.", "gold_files": ["requests/models.py", "requests/exceptions.py"], "anchor_files": [], "budget_tokens": 4000, "description": "raise_for_status enhancement in models and exceptions.", "is_held_out_repo": false}
+{"id": "req_selective_01", "repo_name": "psf/requests", "task_type": "selective", "query": "Optimize PyTorch GPU tensor CUDA backward autograd pass for multi-head attention weights.", "gold_files": [], "anchor_files": [], "budget_tokens": 4000, "description": "Counterfactual PyTorch CUDA query in Requests.", "is_held_out_repo": false}

schema.json

@@ -0,0 +1,50 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "title": "RepoNeedleBenchSample",
+  "description": "Schema for RepoNeedle-Bench evaluation instances.",
+  "type": "object",
+  "properties": {
+    "id": {
+      "type": "string",
+      "description": "Unique sample identifier."
+    },
+    "repo_name": {
+      "type": "string",
+      "description": "Repository origin identifier (e.g. pallets/flask, psf/requests, tiangolo/fastapi)."
+    },
+    "task_type": {
+      "type": "string",
+      "enum": ["code2test", "comment2context", "trace2code", "edit2ripple", "selective"],
+      "description": "Workflow retrieval category."
+    },
+    "query": {
+      "type": "string",
+      "description": "Input query, issue comment, traceback, or code fragment."
+    },
+    "gold_files": {
+      "type": "array",
+      "items": { "type": "string" },
+      "description": "Ground truth repository file paths required to resolve the task."
+    },
+    "anchor_files": {
+      "type": "array",
+      "items": { "type": "string" },
+      "description": "Anchor or modified source files for ripple/impact tasks."
+    },
+    "budget_tokens": {
+      "type": "integer",
+      "default": 4000,
+      "description": "Context token budget constraint."
+    },
+    "description": {
+      "type": "string",
+      "description": "Human-readable task summary."
+    },
+    "is_held_out_repo": {
+      "type": "boolean",
+      "description": "Whether the repository is part of the held-out evaluation split."
+    }
+  },
+  "required": ["id", "repo_name", "task_type", "query", "gold_files", "budget_tokens"]
+}
+