chore: remove hardcoded repo paths and add community docs

Author: sjarmak

CODE_OF_CONDUCT.md

@@ -0,0 +1,20 @@
+# Code of Conduct
+
+## Our Standards
+
+Contributors are expected to be respectful, constructive, and professional in issues, pull requests, and reviews.
+
+Unacceptable behavior includes:
+
+- Harassment, discrimination, or personal attacks
+- Trolling, insulting, or derogatory comments
+- Publishing private information without permission
+- Disruptive or bad-faith participation
+
+## Enforcement
+
+Project maintainers may remove, edit, or reject comments and contributions that violate this code of conduct.
+
+## Reporting
+
+If you experience or witness unacceptable behavior, report it privately to the maintainers through GitHub (do not open a public issue for conduct reports).

CONTRIBUTING.md

@@ -0,0 +1,40 @@
+# Contributing
+
+Thanks for contributing to CodeContextBench.
+
+## Scope
+
+This repository contains benchmark tasks, evaluation configs, and analysis tooling. Changes should preserve reproducibility and avoid breaking existing task definitions unless the change is intentional and documented.
+
+## Before Opening a PR
+
+Run the repo health gate from the repository root:
+
+```bash
+python3 scripts/repo_health.py
+```
+
+For docs/config-only changes, `--quick` is usually sufficient during iteration:
+
+```bash
+python3 scripts/repo_health.py --quick
+```
+
+## Change Guidelines
+
+- Keep changes scoped and easy to review.
+- Update documentation when behavior, workflows, or file locations change.
+- Avoid committing local run artifacts (`runs/`, `results/`, `eval_reports/`, logs).
+- Do not commit credentials, tokens, or private keys.
+
+## Benchmark / Task Changes
+
+- Validate tasks before launching runs: `python3 scripts/validate_tasks_preflight.py --all`
+- Reconcile task metadata when needed: `python3 scripts/sync_task_metadata.py --help`
+- Follow task-oriented docs in `docs/START_HERE_BY_TASK.md`
+
+## Pull Requests
+
+- Explain what changed and why.
+- Call out any benchmark-impacting behavior changes.
+- Include validation commands run (health gate, tests, or scripts) in the PR description.

SECURITY.md

@@ -0,0 +1,21 @@
+# Security Policy
+
+## Reporting a Vulnerability
+
+Please do not open a public GitHub issue for suspected security vulnerabilities.
+
+Instead:
+
+- Use GitHub Security Advisories / private vulnerability reporting for this repository if available.
+- If private reporting is not available, contact the maintainers privately through GitHub.
+
+Include:
+
+- A clear description of the issue
+- Steps to reproduce
+- Impact assessment (what could happen)
+- Any proposed mitigation or patch (if available)
+
+## Scope Notes
+
+This repository contains benchmark tasks and tooling and may include third-party fixture code under benchmark environments. If the issue is in third-party upstream code, please also report it to the upstream project maintainers when appropriate.

scripts/analyze_mcp_unique_haiku.py

@@ -24,7 +24,7 @@
 
 # -- Configuration --
 
-STAGING_DIR = Path("/home/stephanie_jarmak/CodeContextBench/runs/staging")
+STAGING_DIR = Path(__file__).resolve().parent.parent / "runs" / "staging"
 RUN_SUFFIX = "_haiku_20260221_140913"
 BASELINE_CONFIG = "baseline-local-artifact"
 MCP_CONFIG = "mcp-remote-artifact"

scripts/analyze_run_coverage.py

@@ -12,7 +12,7 @@
 from collections import defaultdict
 from pathlib import Path
 
-BASE_DIR = Path("/home/stephanie_jarmak/CodeContextBench")
+BASE_DIR = Path(__file__).resolve().parent.parent
 SELECTED_TASKS_FILE = BASE_DIR / "configs" / "selected_benchmark_tasks.json"
 RUNS_DIR = BASE_DIR / "runs" / "official"
 

scripts/backfill_instruction_artifacts.py

@@ -178,9 +178,10 @@ def find_instruction(task_path: str, trial_dir: str) -> str:
 def main():
     dry_run = "--dry-run" in sys.argv
 
+    repo_root = Path(__file__).resolve().parent.parent
     search_roots = [
-        Path(__file__).resolve().parent.parent / "runs" / "official",
-        Path("/home/stephanie_jarmak/CodeContextBench/runs/validation"),
+        repo_root / "runs" / "official",
+        repo_root / "runs" / "validation",
     ]
 
     trial_count = 0

scripts/consolidate_staging.py

@@ -22,14 +22,15 @@
 from collections import defaultdict
 from pathlib import Path
 
-STAGING = Path("/home/stephanie_jarmak/CodeContextBench/runs/staging")
+REPO_ROOT = Path(__file__).resolve().parent.parent
+STAGING = REPO_ROOT / "runs" / "staging"
 
 
 def get_aggregate_data():
     result = subprocess.run(
         ["python3", "scripts/aggregate_status.py", "--staging"],
         capture_output=True, text=True,
-        cwd="/home/stephanie_jarmak/CodeContextBench",
+        cwd=str(REPO_ROOT),
     )
     return json.loads(result.stdout)
 

scripts/find_mcp_distracted.py

@@ -15,7 +15,7 @@
 from pathlib import Path
 from collections import defaultdict
 
-STAGING_DIR = Path("/home/stephanie_jarmak/CodeContextBench/runs/staging")
+STAGING_DIR = Path(__file__).resolve().parent.parent / "runs" / "staging"
 
 SDLC_SUITES = {"build", "debug", "design", "document", "fix", "secure", "test", "understand"}
 

scripts/judge_demo.py

@@ -53,17 +53,19 @@
 # Config
 # ──────────────────────────────────────────────────────────────────────
 
+REPO_ROOT = Path(__file__).resolve().parent.parent
+
 BL_TASK_DIR = Path(
-    "/home/stephanie_jarmak/CodeContextBench/runs/staging/"
-    "test_sonnet_20260219_015458_to_20260219_213932/"
-    "baseline/ccb_test_test-unitgen-go-001_baseline/"
+    REPO_ROOT / "runs" / "staging" /
+    "test_sonnet_20260219_015458_to_20260219_213932" /
+    "baseline" / "ccb_test_test-unitgen-go-001_baseline" /
     "test-unitgen-go-001__7SxXhsF"
 )
 
 SG_TASK_DIR = Path(
-    "/home/stephanie_jarmak/CodeContextBench/runs/staging/"
-    "test_sonnet_20260219_015458_to_20260219_213932/"
-    "sourcegraph_full/ccb_test_test-unitgen-go-001_sourcegraph_full/"
+    REPO_ROOT / "runs" / "staging" /
+    "test_sonnet_20260219_015458_to_20260219_213932" /
+    "sourcegraph_full" / "ccb_test_test-unitgen-go-001_sourcegraph_full" /
     "sdlc_test_test-unitgen-go-001_TJ__VwAEJTL"
 )
 

scripts/monitor_and_queue.sh

@@ -5,7 +5,8 @@ set -euo pipefail
 # then automatically start SF runs when each finishes.
 
 POLL_INTERVAL=60
-WORKDIR="/home/stephanie_jarmak/CodeContextBench"
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+WORKDIR="$(cd "${SCRIPT_DIR}/.." && pwd)"
 
 PYTORCH_SB_SESSION="pytorch"
 SWEBENCH_SB_SESSION="swebench-base"