Source-backed example. Practical decision guidance is proposed and untested; this is not a full implementation guide.
The problem
Writing assistance can feel inconsistent when it lacks examples of how someone actually communicates and forgets their deliberate style rules.
Agent job
Read approved writing samples, create a personal voice skill and merge later patterns while preserving manual rules.
Expected output
A reviewable writing-style skill grounded in supplied or authorized sent-message examples.
Where this framing comes from
Core problem: Editorial purpose framing
Agent job: Editorial purpose framing
Expected output: Editorial purpose framing
Relevant job roles: Editorial purpose framing
Benefit type: Editorial purpose framing
Editorial framing and outcome limits
Describes the task’s purpose and intended output; measured benefit, safe unsupervised operation and successful delivery are not established by this framing.
Editorial synthesis based on the linked source. This framing does not add evidence of successful delivery or results.
Case-specific decision guidance · untested
Case-specific decision guidance and first checks are original proposals. The linked source supports the attributed task; this added guidance has not been executed.
A person may repeatedly correct drafts that sound unlike them, even after giving a few tone instructions. Their concise work updates and warmer personal messages can also look inconsistent when examples are mixed without recipient context.
This job could turn approved examples into a reviewable set of writing tendencies while preserving deliberate rules. The useful artifact explains where each tendency applies and treats weak or conflicting patterns as questions rather than permanent preferences.
Why an agent / simpler alternative
Useful judgment
Infer context-specific patterns, distinguish the user's text from quotations and reconcile new observations with manual instructions.
Simpler alternative
Write a short style checklist when preferences are explicit and stable.
Unsuitable when
Samples contain third-party secrets, ambiguous authorship or too little writing to support a useful pattern.
Proposed inputs
Name
Writing sample pack
Owner
Writer
Minimum scope
Approved excerpts grouped by channel and audience
Quality check
Remove quoted replies and confirm authorship
Name
Manual style rules
Owner
Writer
Minimum scope
Explicit preferences and protected exceptions
Quality check
Mark rules that observations must never overwrite
Access and setup — proposed or unknown
Proposed minimum
Start with supplied excerpts and a separate draft skill; no broad mailbox or workspace access is necessary.
Compatibility unknowns
Current plugin access, skill format, persistence behavior and exact runtime compatibility are unknown.
External actions requiring approval
Reading additional private sources, replacing a shared skill or sending generated messages requires authorization.
Expected deliverable and review checks
Proposed artifact
Draft voice guide with contextual tendencies, supporting sample pointers and a proposed change log.
Critical checks
Preserve manual rules; separate channels; keep personal facts out of general style instructions.
Missing output evidence
No inspected style artifact or validated voice-matching performance is supplied in this overlay.
Failure signals and proposed responses
Signal
A quoted sender's wording becomes a style rule
Why it matters
The guide attributes someone else's voice to the writer
Proposed response
Remove the sample and reconsider dependent rules.
Signal
A one-off apology becomes a lasting preference
Why it matters
Rare circumstances can distort future drafts
Proposed response
Mark the pattern provisional until supported elsewhere.
Effort and cost — unknown
Setup work
Select representative samples and define rule precedence.
Ongoing review
Review proposed changes and remove obsolete contextual patterns.
Unknowns
Sample sufficiency, maintenance frequency, privacy review and service costs are unknown.
First check — proposed, unexecuted
Status
proposed_unexecuted
Synthetic fixture
Invent several short messages across two audiences, one quoted reply and a manual rule contradicting an apparent pattern.
Acceptance checks
Exclude the quotation; preserve the manual rule; qualify audience-specific patterns; make no persistent update.
Stop conditions
Stop when authorship is uncertain or inference would require unauthorized private history.
Limits
A style guide should not invent beliefs, commitments or relationship details on the writer's behalf.
Editorial relevance from the title and source-supported actions; not evidence of the operator’s occupation.
Claude CoworkIndependent testing: No
Editorial candidate. This is source-backed reporting or a documented workflow. BotCurrent has not independently tested this task. A proposed Canadian adaptation is not evidence of a Canadian deployment or compliance certification.