LYGO Ops Detector

LYGO Ops Detector — local AETHONΔ9 discourse heuristics for evasion, half-truth certainty, saturation bait, coordination language, and policy-refusal signals in operator-supplied text, plus weighted public metadata (account_based_in mismatch, VPN-hint, HTTPS-cited incidents, same-geo+scripted batch). Opt-in only. Stdlib CLI; file reads require --i-consent; eval writes under tests/. Not for doxing or nationality guilt. Dual-threshold (operational 0.65 vs calibration). Pairs with lygo-flame-ward. Triggers: lygo ops detector, aethon d9, evasion index (explicit).

LYGO Ops Detector — AETHONΔ9 v1.4.0

Local, deterministic discourse-signal heuristics plus weighted public context.
Not a person profiler. Not sole evidence. Not for doxing.
Country labels are not guilt. A geo field alone cannot clear the 0.65 bar.

Signature: Delta9Phi963-OPS-DETECTOR-v1.4.0
Pairs with: lygo-flame-ward (authority gate) · lygo-deception-radar (public proof)

---

What's new in 1.4.x

Add	Why
--public-meta / --public-meta-file	Operator-supplied public fields are always scored when present
Weighted public context	Geo mismatch, location_accurate=false, HTTPS-cited incident, same-geo+scripted batch
Hard gate	Country-only / nationality cannot clear ops_score 0.65
No country denylist	Nigeria, India, USA, etc. score the same as a lone label
Flame hints	public_meta_mismatch · named_public_incident

1.3.x still applies: half-truth, saturation, flame hints, cluster boost, 0.65 bar.

---

Permissions

Capability	Default
Network	None
Shell / subprocess	None
Read local files	--text-file / --assoc-file / --public-meta-file + --i-consent
Write	Eval report under tests/ only
Auto-publish / social	Never

When to invoke (narrow)

Only when the user explicitly wants ops-detector / AETHONΔ9 / evasion index on text they supply.

Do not auto-trigger on generic “analyze this email.”

What it measures

Channel	Measures	Does not
Evasion	Burden-shift, ad hominem, vague cites, authority inflation, gaslight, deflection	Person identity
Half-truth certainty	Prestige/certainty shut-downs without primary digests	Medical/legal truth claims
Saturation bait	Rage/click attention-weapon templates	Ordinary urgency with cites
Association	Coordination/secrecy language you provide	Social-graph doxing
Institutional	Policy-as-shield / no-comment	Affiliation / faith / job titles
Public context	Operator-supplied public fields: account_based_in vs claimed location, VPN/inaccurate flag, HTTPS-cited public incident, same-geo+scripted batch	Nationality guilt, live X scrape, country denylist

Thresholds (honest)

Bar	Meaning
Operational ops_score >= 0.65 or evasion_index >= 0.65	Strong multi-signal bar for human review
Calibration (low)	Suite ranking only — not production marketing

Safe use

bash

cd path/to/lygo-ops-detector
python scripts/self_check.py
python scripts/lygo_ops_detector.py --text "paste discourse here" --json
python scripts/lygo_ops_detector.py --text-file ./snippet.txt --i-consent
python scripts/lygo_ops_detector.py --text "I'm based in the United States. ..." --public-meta "{\"account_based_in\":\"Nigeria\",\"claimed_location\":\"United States\",\"location_accurate\":false}" --json
python scripts/lygo_ops_detector.py --show-boundaries

Exit: 0 clear/low · 3 need consent · 10 high evasion discourse (review claims).

Flame bridge

JSON field flame_enemy_hints may include half_truth_pack · authority_shield · saturation_flood · public_meta_mismatch · named_public_incident.
Then run Flame ingest-gate before crowning lattice authority:

bash

python path/to/lygo-flame-ward/scripts/flame_cli.py ingest-gate --text "..."

Agent contract

1. Call the script for reproducible scores.
2. Separate observed regex hits vs inference.
3. Never name people as investigation targets from this tool alone.
4. Discourse pattern ≠ guilt. Country label ≠ guilt.
5. Public metadata must be operator-supplied (no live scrape). HTTPS URL required for named-incident weight.
6. No external publish without user consent.
7. File inputs need --i-consent.

Security

Read references/SECURITY.md + references/SKILLSPECTOR_AUDIT.md.

Ver	Change
1.2.2	SkillSpector discourse-not-identity harden
1.3.0	Half-truth + saturation · Flame hints · package polish
1.3.1	Cluster boost · evasion bar 0.65 · suite multi-signal · operational metrics fixed
1.4.0	Weighted public metadata (mismatch / VPN-hint / HTTPS incident / geo+template). No country denylist. Country-only cannot clear 0.65.

Δ9Φ963 — receipts over hype · discourse not identity · seals first with Flame.