Retribution Accountability Tracker — Query Skill
Last updated: September 16, 2026
What Data Is Available
The Retribution Accountability Tracker is a validated JSON ledger — agent-base/tracker/data/retribution_events.json (schema: retribution_events.schema.json) — governed by docs/retribution-methodology.md. One record per retributive act. A rendered, reader-facing view lives at the KB doc truth-reconciliation/retribution-accountability-tracker.md, and a Flask dashboard at /retribution.
Each event carries:
target—{name, type, profile_slug?}(individual / company / university / law_firm / media / federal_employee / organization / official)act_type— one of:firing_removal,prosecution,subpoena,investigation,clearance_revocation,contract_cancellation,funding_freeze,license_bar_targeting,immigration_weaponization,access_revocation,regulatory_pressure,executive_orderresponsible[]— the responsibility chain: each party has arole(orderer / approver / signatory / notifier / executor / spokesperson), anevidence_tier, aconfidence, and aresponsibilityflag (documented→ may be named;inferred→ role only, never named)rss— Retribution Severity Score 1–5harm{}— career / economic_security / reputation / family / chilling_effect (none/low/moderate/severe)cruelty_evidence[]— sourced on-the-record quotes evidencing punitive intenttrc_classification— A (prosecution track) / B (institutional-reform track) / C (truth-telling / record)status,confidence,sources[],dms_ref(link to a profile’s Democratic Malice Assessment)
The Naming Rule (read before answering)
A responsible party is reported as a named individual ONLY when responsibility: documented. Where responsibility is inferred, report the role and agency only (“a DOJ signatory,” “the OMB director’s office”) — never assert a named person as the actor. When an answer surfaces inferred attribution, say so explicitly (“this attribution is inferred, not documented”). This mirrors the defamation posture in docs/retribution-methodology.md §3 and .cursor/rules/patriot-retribution-tracker.mdc.
How to Query
By target
Find every act against a target by searching events[].target.name. Example: “What was done to Harvard?” → the funding_freeze event(s) with target.name: Harvard University, their RSS, harm, and (documented) responsible officials.
By official (the accountability question)
The core use. To answer “who is responsible for the most / most severe retribution,” read the responsibility leaderboard — aggregates.responsibility_leaderboard, ranked by severity-weighted score (Σ RSS across documented attributions) then act count. Only documented, named parties appear there. For one official: RetributionTracker.acts_for_official(slug_or_name).
By act type
Group by act_type (aggregates.by_act_type). Example: “Which law firms were hit with executive orders?” → act_type: executive_order with target.type: law_firm.
By severity / TRC track
Filter rss >= 4 for the gravest acts, or trc_classification: A for acts with plausible criminal exposure on the prosecution track.
Cross-Reference with Accountability Profiles
When a documented responsible party carries a profile_slug (or dms_ref), cross-link to their accountability profile’s Retribution Actions section and Democratic Malice Assessment. Officials without a profile are enumerated in docs/retribution-profile-backlog.md (the profile-gap list), ranked to sequence profile creation.
Updating the Data
The ledger is populated by scripts/ingest_retribution_events.py (LLM extraction from the narrative corpus + attribution research) and rendered to the KB doc by scripts/render_retribution_tracker.py. The graph layer is built by scripts/build_retribution_graph.py → docs/retribution-graph.json. Publishing stays human-gated — never run make publish autonomously.
Composing Skills
This skill composes with:
retribution-assessment-framework— score a new act (gate + RSS + harm + TRC)retribution-responsibility-tracer— attribute an act to officials to the required evidence tiermalice-evaluator— Democratic Malice Assessment for a named perpetrator’s profilepublic-corruption-ombudsman— officials who weaponize officelegal-research-specialist— court-filing verification of prosecutions/subpoenas
