What can I use SignalFlow for today?

SignalFlow can capture manual ContentSignals, connect verified GitHub repository context into hosted Signals and Opportunities, continue connected Opportunities into durable Voice and NarrativeStrategy planning, and run the accepted exact-review path for LinkedIn and X. The legacy campaign builder remains compatibility-only while canonical Direct Create is rebuilt on the same Content OS records.

SignalFlow is an approval-first content operating system. It is designed to reduce the operational work between meaningful work and audience communication while keeping judgment, explicit boundaries, and exact approval with the user.

Does SignalFlow publish without approval?

No. Approval is bound to an exact visible revision. A changed or regenerated draft requires judgment again, and publication is kept separate from preparation until the external outcome is confirmed.

Where is current state stored?

The accepted personal owner path still uses browser-local recovery for some records, while GitHub source connections, ProjectContext, Signals, Opportunities, Identity, NarrativeStrategy, and ContentPiece continuity are being moved into the hosted relational path. SignalFlow does not describe prepared content as publicly published until an external publication is confirmed.

Can I bring my own model provider?

SignalFlow has a provider-neutral inference boundary and supports configured hosted, bring-your-own, local, and compatible custom routes where the active deployment can reach them. Privacy policy is part of routing and protected context must fail closed instead of silently falling back to a weaker route.

VOICE · EXPLICIT IDENTITY

Teach SignalFlow what should remain recognizably you.

This is not a tone preset. These are explicit, versioned instructions about identity, perception, voice and boundaries. Learned preferences sit below them and remain visible and controllable.

Identity Not setVoice Not setBoundaries Not set

01 · IDENTITY

What do you actually work on and care about?

Enough context to prevent SignalFlow from reducing you to one job title or one project.

Primary things you work / talk about

Expertise / credible contextInterests beyond the obvious

Background worth knowing when relevant

02 · PERCEPTION

How should people understand you over time?

Desired perception is different from sounding impressive in every post.

What should a thoughtful reader come away believing?

Qualities to signalQualities / impressions to avoid

03 · VOICE

What makes AI-written content stop sounding like you?

Write rules in plain language. Explicit rules always outrank learned preferences.

Writing principlesThings you dislike in AI writing

Optional example of your own writing

Technical depthPersonal vulnerabilityEmoji

04 · BOUNDARIES

What must SignalFlow never invent, expose or turn into a style?

Explicit boundaries have higher precedence than campaign instructions, platform conventions, or learned preferences.

Blocked phrases / claimsOther hard rules

Built in for this profile: unverified metrics and fabricated vulnerability are block-level rules; exaggerated launch language is flagged by default.

05 · PLATFORM EXPRESSION

Same person. Different native expression.

LinkedIn and X can adapt depth and pacing, but neither may override global boundaries.

LinkedInX

06 · LEARNED PREFERENCES

What SignalFlow has inferred from your real corrections.

These are explainable hypotheses, not a hidden personality profile. Candidate patterns need evidence; confirmed patterns still remain below explicit Voice and boundaries.

0 learned preferencesReset removes style hypotheses only. It does not delete campaign, approval, or publication history.

No learned preferences yet. SignalFlow will only create them from eligible review feedback; one isolated edit normally remains a candidate rather than becoming a rule.

Save explicit profileCreates new versions; it does not erase the previous profile history.

VOICE · EXPLICIT IDENTITY

Teach SignalFlow what should remain recognizably you.

This is not a tone preset. These are explicit, versioned instructions about identity, perception, voice and boundaries. Learned preferences sit below them and remain visible and controllable.

Identity Not setVoice Not setBoundaries Not set

01 · IDENTITY

What do you actually work on and care about?

Enough context to prevent SignalFlow from reducing you to one job title or one project.

Primary things you work / talk about

Expertise / credible contextInterests beyond the obvious

Background worth knowing when relevant

02 · PERCEPTION

How should people understand you over time?

Desired perception is different from sounding impressive in every post.

What should a thoughtful reader come away believing?

Qualities to signalQualities / impressions to avoid

03 · VOICE

What makes AI-written content stop sounding like you?

Write rules in plain language. Explicit rules always outrank learned preferences.

Writing principlesThings you dislike in AI writing

Optional example of your own writing

Technical depthPersonal vulnerabilityEmoji

04 · BOUNDARIES

What must SignalFlow never invent, expose or turn into a style?

Explicit boundaries have higher precedence than campaign instructions, platform conventions, or learned preferences.

Blocked phrases / claimsOther hard rules

Built in for this profile: unverified metrics and fabricated vulnerability are block-level rules; exaggerated launch language is flagged by default.

05 · PLATFORM EXPRESSION

Same person. Different native expression.

LinkedIn and X can adapt depth and pacing, but neither may override global boundaries.

LinkedInX

06 · LEARNED PREFERENCES

What SignalFlow has inferred from your real corrections.

These are explainable hypotheses, not a hidden personality profile. Candidate patterns need evidence; confirmed patterns still remain below explicit Voice and boundaries.

0 learned preferencesReset removes style hypotheses only. It does not delete campaign, approval, or publication history.

No learned preferences yet. SignalFlow will only create them from eligible review feedback; one isolated edit normally remains a candidate rather than becoming a rule.

Save explicit profileCreates new versions; it does not erase the previous profile history.

What do you actually work on and care about?

How should people understand you over time?

What makes AI-written content stop sounding like you?

What must SignalFlow never invent, expose or turn into a style?

Same person. Different native expression.

What SignalFlow has inferred from your real corrections.

Preparing your workspace

What do you actually work on and care about?

How should people understand you over time?

What makes AI-written content stop sounding like you?

What must SignalFlow never invent, expose or turn into a style?

Same person. Different native expression.

What SignalFlow has inferred from your real corrections.