Frequently Asked Questions

Go to Import Messages and select a WhatsApp, iMessage, or generic CSV/TXT export from your device. The tool accepts exports from iPhone message export tool, iPhone Backup Extractor, WhatsApp's built-in export, and plain text logs.

If you have multiple files (e.g., monthly WhatsApp exports), you can import them one at a time — the tool deduplicates automatically so you won't end up with duplicate messages.

The tool currently supports:

• WhatsApp exports (.txt) — the standard "Export Chat" format from WhatsApp settings
• iMessage / SMS exports from iPhone message export tool or iPhone Backup Extractor (CSV or TXT)
• Generic CSV with timestamp, sender, and content columns

If your export format isn't working, try the generic CSV option — most messaging apps can export in this format.

When you import messages, senders are identified by name as they appear in your export file (e.g. "John", "+447911123456"). Sender Identities let you map these raw names to roles: Me, Other Parent, or Third Party.

This matters because the analysis focuses on messages from the other parent. Getting this mapping right ensures the metrics and flagging are accurate.

Yes — analysis is not automatic. After importing new messages, go to Run Analysis and trigger it. You can run it on all conversations or a specific import. Analysis typically takes a few seconds to a couple of minutes depending on message volume.

The tool uses two complementary systems:

1. Keyword & pattern matching — each message is checked against a library of over 400 keywords and phrases organised into categories (Harassment & Threats, Coercive Control, Legal Threats, Parental Alienation, etc.). Matches create flags with a severity from 1 (minor) to 5 (critical).
2. 5-axis classifier — each message is also scored on communication type (child-logistics, emotional, legal, etc.), tone, escalation level, and whether it contains safeguarding concerns.

Both systems run together during analysis. The flags appear in Review Flags; the classifier results drive the Metrics charts.

Go to Review Flags, find the flag, and use the Dismiss action to remove false positives, or Confirm to mark it as verified. You can also add a note explaining the context.

Dismissed flags are hidden from exports and metrics by default but remain in the database — they are not permanently deleted so you can review them later.

The Child-Focused Messages metric counts messages that are primarily about child logistics or welfare — school runs, medical appointments, activity planning, etc. A low percentage means a higher proportion of communication is emotionally charged, conflict-driven, or about adult issues rather than the children's needs.

Family courts consider a parent's ability to keep communication child-focused as a marker of good co-parenting. A persistently low score may be relevant evidence.

Escalation rate measures what proportion of the other parent's messages contain escalatory language — threats, aggression, manipulation, or legal intimidation. It is calculated from the 5-axis classifier results and the flag database combined.

Click the escalation rate card on the Metrics page to drill down to the underlying flagged messages.

The tool produces structured evidence documents — chronologies, flag summaries, and pattern analyses — but this is not legal advice. Admissibility depends on how evidence is presented and whether it is relevant to the specific proceedings.

Generally, message content itself is admissible (it is your own communications). The tool helps you organise and present that content clearly. Discuss with your solicitor how best to use the output.

Go to Generate Documents. The documents available depend on your journey (C100 Applicant, C7 Respondent, Parenting Plan, Consent Order), and are grouped by relevance:

• Position Statement — structured statement of your concerns, cross-referenced to flags and the welfare checklist
• Evidence Timeline — date-ordered incident chronology in standard UK family court bundle format
• Welfare Checklist Report — s.1(3) Children Act 1989 factors addressed with your evidence
• Positive Parenting Evidence Pack — evidence of your child-focused conduct, organised by welfare factor
• Scott Schedule — fact-finding format (DA/CC allegations), required under PD12J
• Parenting Plan Agreement — CAFCASS-framework co-parenting plan, suitable for co-signing or filing with a Consent Order
• Consent Order Draft — draft consent order in standard UK family court format (Recitals + numbered paragraphs)
• C100 / C7 Form Notes wizards — guided draft-answer documents for completing the official HMCTS C100 (application) and C7 (acknowledgement) forms

All documents are marked as drafts. Review with a solicitor or McKenzie Friend before filing.

You can control this in Settings → Include my own flags in exports. Including your own messages provides context and demonstrates transparency, but may also highlight messages that could be used against you. Discuss with your solicitor.

By default, your own flagged messages are included in analysis (so you can see the full picture) but excluded from exports.

The tool does not add authentication — message authenticity is established by the export source (WhatsApp, iMessage, etc.). Retain original export files and be prepared to explain how you exported them. A phone forensics expert can verify authenticity if challenged.

This is a deliberate design choice — not just a technical constraint. Family court evidence is among the most sensitive personal data there is: it includes details of domestic abuse allegations, children's welfare, and communications that could affect care arrangements.

Storing data locally means:

• No breach risk. Your data cannot be accessed by a server hack, a vendor going bust, or a subpoena to a third-party provider.
• Full legal clarity. The data is yours, on your machine, under your control — it never becomes someone else's problem to protect.
• No UK GDPR complexity for you. Because no personal data leaves your device, there are no cloud processing agreements, third-party processors, or data residency questions to navigate.

The only exception: if you use the optional AI-enhanced analysis feature, individual flag snippets are sent to OpenAI's API for contextual review. This is opt-in and you control whether it runs.

All data is stored locally on your computer in a SQLite database file inside the tool's folder. Nothing is sent to any server or cloud service. The tool runs entirely offline.

Yes — copy the coparent_tool folder to a backup location. The database file is coparent_tool/coparent.db. Keep this backed up regularly, especially before running large analyses or making significant changes.

Go to Settings → Danger Zone → Reset All Data. This permanently deletes all imported messages, flags, and analysis results. It cannot be undone. Your parenting plan and settings are preserved.