Governance

Policy versions

The violation policy is versioned and immutable. Every decision resolves to the version in force at its own timestamp, so a policy change can never rewrite history. You are acting as operator.

Non-negotiable

Reproducibility guarantee

A policy change never alters a historical decision

Every decision resolves to the version in force at its own timestamp via policyAt(). Publishing supersedes a version — it never edits it — and rollback re-publishes an earlier version rather than deleting the one it replaces.

Versions on record

Each published version is immutable.

Resolves now

v1.0.0

Version in force at this instant.

Earliest in-force

v1.0.0

At 2026-09-29 09:24 UTC it still resolves to itself.

Demonstrated against the local store: a decision recorded at 2026-09-29 09:24 UTC resolves to version 1.0.0 — the version published then, not today's active version.

Mutating operations

Policy actions

Read-only for your role

You are acting as operator. Drafting, submitting and publishing require the reviewer or admin role; rollback requires admin. The API enforces the same check.

Target versionThe version the next action applies to.

Rollback reasonRe-publishes the selected version as the one in force. Admin only; minimum 10 characters.

Forking copies the 5 rules of active version 1.0.0 into a new draft. Historical decisions remain pinned to their own version.

Every action is recorded in the audit log and roles are re-checked server-side. Nothing here can approve on an operator's behalf.

Protected expression

Express non-grounds

Criticism of government
Criticism of government, public officials, or policy — on any grounds — is never a violation.
Religious identity
Religious identity, belief, or theological discussion alone is never a basis for action.
Satire
Satire and parody, including sharp political satire, are protected expression.
Reporting
Reporting on a threat, scam, or abuse is not itself the violation.
Quotation
Quoting offending words in commentary or analysis does not endorse them.
Counterspeech
Repeating offending words to condemn or rebut them is expressly permitted.

These categories are reinforced per-rule in each rule's exceptions list below.

Rules in force now

Active policy — version 1.0.0

Published

Initial published policy set. Five objective violation categories; political criticism, religious identity, satire, reporting, quotation and counterspeech are express non-grounds.

PC-01

Credible threat of violence

Credible threathighReviewer + admin

A specific, credible threat of violence against an identifiable person or group, with sufficient detail to suggest intent and capability.

Evidence required

verbatim passage
timestamped capture
context assessment

Exceptions — never grounds on their own

reporting on a threat made elsewhere
quotation of a threat in commentary
documentation of abuse received

Violation — meets threshold

“Public post naming a specific individual and stating when and where they will be attacked.”
Specific target + time + place = actionable.

Not a violation — does NOT meet threshold

“Journalist quoting a threatening message they received.”
Reporting/quotation, not the journalist's threat.
“Satirical post expressing frustration with a public figure.”
Political criticism is expressly not a ground.

PC-02

Incitement to violence

Incitement to violencehighReviewer + admin

Content urging others to commit violence against a group or individual, where the urging is directed and likely to be understood as such.

Evidence required

verbatim passage
language and translation
context assessment
audience reach note

Exceptions — never grounds on their own

counterspeech that repeats the inciting words in order to condemn them
academic or legal analysis

Violation — meets threshold

“Post calling for a crowd to attack a named community at a stated place.”
Directed urging to violence.

Not a violation — does NOT meet threshold

“A condemnation that quotes the incitement verbatim.”
Counterspeech carries the words without endorsing them.

PC-03

Targeted harassment

Targeted harassmentmediumReviewer approval

Sustained targeting of an identifiable individual with abusive content, including coordinated pile-ons and publishing private information.

Evidence required

verbatim passages
pattern over time
target identified

Exceptions — never grounds on their own

public-interest criticism of a public figure's conduct
satire
one-off disagreement

Violation — meets threshold

“Series of posts over days publishing an individual's address and phone number.”
Privacy violation + pattern.

Not a violation — does NOT meet threshold

“Critical commentary on a minister's policy.”
Criticism of government alone must never trigger action.

PC-04

Fraud and financial deception

Fraud & financial deceptionmediumReviewer approval

Content promoting a scheme intended to deceive people out of money, including fake investment, job and loan offers.

Evidence required

verbatim passage
linked destination
payment or contact method

Exceptions — never grounds on their own

reports WARNING about a scam
regulatory advisories

Violation — meets threshold

“Post advertising guaranteed 5x returns via a WhatsApp link.”
Deceptive financial promotion.

Not a violation — does NOT meet threshold

“Screenshots shared to warn the public about a scam.”
A warning is not the fraud.

PC-05

Hate speech against a protected characteristic

Hate speechhighReviewer + admin

Content that dehumanises or calls for harm against people on the basis of a protected characteristic.

Evidence required

verbatim passage
translation
context assessment

Exceptions — never grounds on their own

criticism of ideas, policies or governments
religious discussion
reporting
counterspeech

Violation — meets threshold

“Post calling for the expulsion of a named community.”
Targets a protected group.

Not a violation — does NOT meet threshold

“Criticism of a government's policy on any grounds.”
Political criticism is expressly excluded.
“Theological argument between denominations.”
Religious identity alone is never a basis.

Every version, immutable

Version timeline

v1.0.0PublishedIn force now
Initial published policy set. Five objective violation categories; political criticism, religious identity, satire, reporting, quotation and counterspeech are express non-grounds.
Published: 2026-09-29 09:24 UTC by M. F. AshrafCreated: 2026-09-29 09:24 UTC by M. F. AshrafIn force: 2026-09-29 09:24 UTC → presentRules: 5