Code of Conduct

Status: v0.1 Public Draft

The LockedIn FDE Benchmark brings together people with different roles, disciplines, organizations, identities, and views of professional capability. Participation requires rigor without hostility and disagreement without dehumanization.

1. Expected behavior

Participants must:

• critique ideas, methods, evidence, and effects rather than attacking people;
• communicate clearly, respectfully, and in good faith;
• make room for different levels of expertise and different demonstration modes;
• disclose material conflicts and correct mistakes promptly;
• protect private, confidential, security-sensitive, and embargoed information;
• avoid implying that contribution equals endorsement, partnership, or verified capability;
• respect accessibility needs, names, pronouns, cultures, and communication differences;
• stop unwanted contact and follow moderator direction;
• raise measurement, bias, safety, or integrity concerns without retaliation.

2. Unacceptable behavior

The following is prohibited in project spaces:

• harassment, threats, stalking, intimidation, or discriminatory conduct;
• sexualized attention or content directed at another participant;
• insults, demeaning labels, dogpiling, or deliberate humiliation;
• publishing or threatening to publish private information;
• submitting private candidate/customer evidence or identifiable allegations publicly;
• impersonation, fabricated authority, undisclosed coordinated influence, or reference collusion;
• attempts to obtain or disclose hidden cases, credentials, private prompts, or security controls;
• retaliation against a reporter, witness, reviewer, candidate, or dissenter;
• repeated disruption after a moderator has set a reasonable boundary.

Good-faith disagreement with the benchmark, LockedIn Labs, or a decision is not misconduct.

3. Scope

This policy applies in repositories, issues, discussions, RFCs, reviews, working groups, project events, evaluation interactions, direct project communications, and public representation made in an official project role.

Private behavior may be in scope when it creates a credible safety risk or materially affects a person’s ability to participate in the project.

4. Reporting

Never report conduct involving personal information through a public issue.

Before managed public participation opens, the project must publish a private conduct-reporting channel and name the people authorized to receive reports in LIVE_VS_PLANNED.md. If no private channel is listed, the community program is not yet operational and no sensitive report should be submitted through this repository.

A report should include, where safe:

• what occurred and where;
• relevant dates and links;
• immediate safety concerns;
• the requested communication method;
• supporting material the reporter has the right to share.

Anonymous reports may be considered, though limited information can constrain investigation and remedy.

5. Response principles

Authorized responders must:

• acknowledge the report without promising a predetermined outcome;
• protect confidentiality to the extent possible and lawful;
• disclose and resolve conflicts before investigating;
• separate the respondent from decisions about their own case;
• avoid unnecessary collection or retention of sensitive information;
• give affected people a fair opportunity to be heard where safe;
• document findings, reasons, actions, and retention;
• prohibit retaliation.

Immediate, narrowly scoped protective action may be taken before a final finding when safety or evidence preservation requires it.

6. Proportionate outcomes

Depending on severity, pattern, context, and risk, outcomes may include:

1. Private clarification or corrective guidance.
2. Formal warning and required behavior change.
3. Temporary restriction from a channel, review, or event.
4. Removal from a role or decision.
5. Time-limited or permanent participation ban.
6. Referral to an employer, platform, venue, or authority where necessary and lawful.

An outcome must not alter a benchmark result as punishment. Evidence or integrity issues discovered through a conduct report must follow the applicable benchmark review and appeal process.

7. Review and appeal

A person subject to a material conduct outcome may request review by a non-conflicted person who did not make the original decision. A request does not automatically suspend a protective action.

The project may limit details disclosed to protect reporters, witnesses, confidential information, or safety. It should still provide the affected person a meaningful reason and review path.

8. Transparency

The project should publish aggregate conduct reporting without identifying participants: report volume, broad category, response time, outcome category, and material policy improvements.

No claim of an operational conduct program may be made until the private channel, responsible people, retention rules, and review path are active.