completeFreeSkillFile
ChatGPTVerify uploaded bytes and seal one immutable file. Does not publish or approve content.
configureFreeSkillImprovement
ChatGPTAfter explicit permission, prepare review-only requests for this existing measurement. Reuse an owner-created tactic or create one canonical tactic. No server AI, publication, spending or paid orchestration.
createFreeSkillProject
ChatGPTLegacy retry only. Free skills are fixed to the first selected project; creating another project is refused.
deleteFreeSkillProjectData
ChatGPTDelete this user’s free-skill work in the selected project only when explicitly requested. Preserve unrelated paid data.
downloadFreeSkillFile
ChatGPTGet a short-lived download URL for a saved file.
exportFreeSkillProject
ChatGPTExport one bounded page of personal work and metadata. Follow nextCursor until null. File bytes use the download tool.
finalizeFreeSkillArtifact
ChatGPTCommit a revision only after all files are verified. Confirm the returned saved status.
getFreeSkillArtifact
ChatGPTRead exact revision status and manifest. Set upload=true only to resume an unfinished upload.
getFreeSkillContext
ChatGPTRead user-provided premises and saved work. No formal paid strategy is included.
getFreeSkillDeploymentTimeline
ChatGPTRead paginated application reports and corrections for one exact project and URL, including after pilot stop.
getFreeSkillImprovementData
ChatGPTRead scoped immutable observations and context. Provider-derived; project use only.
getFreeSkillImprovementHandoff
ChatGPTRead the exact child tactic approval saved by the user in MyMarketer. Use its saved source revision and tacticId to start the child with startFreeSkillWork. Do not ask again about its approved brief. Publication, sending and paid operations remain separate.
getFreeSkillImprovementRequest
ChatGPTRead the owner-only automatically prepared review. Host execution is required. Read exact observation IDs via getFreeSkillImprovementData and return analysis using reviewRequestId; no automatic publication or paid run.
getFreeSkillInstructions
ChatGPTRead current server-owned skill procedures for the selected free project. Fetch the manifest, then every required document page at its revision/hash. No server AI, production registration, paid workflow or publication permission.
getFreeSkillMeasurement
ChatGPTRead the selected project measurement setting and failures. Does not collect external data.
getSharedWorkArtifact
ChatGPTRead an exact explicitly shared saved work revision or its file download. Never changes source ownership.
interruptFreeSkillWork
ChatGPTMark an explicitly stopped registered task as aborted.
listFreeSkillProjects
ChatGPTList only projects accessible through this connection.
prepareFreeSkillArtifact
ChatGPTReserve storage and a new immutable revision. Upload is not completed saving. Revisions require the user change request and exact predecessor.
readFreeSkillSource
ChatGPTRead an exact version and bounded unicode-scalar source range from an explicitly selected free project. Raw source only; no server AI or paid context. Metadata classification is host-reported, not approval.
recordFreeSkillAnalysis
ChatGPTRecord host interpretation with exact observation revisions, limitations and next hypotheses. Never causal proof.
recordFreeSkillAssistance
ChatGPTOptionally report human assistance and minutes for this saved revision. Self-report, not independently verified; no research extraction.
recordFreeSkillDeployment
ChatGPTReport a revision application interval or append a correction. This is not verified publication. Requires deployment pilot flag.
recordFreeSkillImprovementCandidates
ChatGPTAfter a requested and saved same-work delivery, save up to three optional next-work candidates with real evidence revision IDs. The user reviews and chooses in MyMarketer. This records suggestions only. Keep the same operationId for retries. For a requested new set use a new operationId.
recordFreeSkillInterpretation
ChatGPTSave host interpretation separately, citing exact observation records. Never establishes causal effects or verified facts.
recordFreeSkillKeyPointCorrection
ChatGPTAppend a correction to one owned key point; never rewrites its original source.
recordFreeSkillKeyPointQuestion
ChatGPTRecord asked, deferred or declined state for a revision key point question. Does not change consent.
recordFreeSkillLifecycle
ChatGPTRecord reported adoption or execution separately from observed results. Host statements are self-reported, not verified publication.
recordFreeSkillObservation
ChatGPTRecord version-specific observation, period, metric definition and provenance. Unknown is not zero.
recordFreeSkillRetrospective
ChatGPTSave a production retrospective, not an observation-based interpretation. Causality is not established. Anchor to the latest saved revision being reviewed; covered revisions must be owned saved versions in the same project. No observation records are required. Does not extract research or append work-record evidence.
reportFreeSkillTransfer
ChatGPTReport a bounded host upload/download outcome for an owned revision. No URLs, raw errors or file content. Self-report only; never confirms saving. Reporting failure must not stop recovery.
startFreeSkillMeasurement
ChatGPTStart the separately permitted one-project GA4 pilot after explicit user configuration. OAuth and research consent alone are not measurement permission. Uses exact property/LP, selected key-event metrics and explicitly permitted lookback (maximum 28 days).
startFreeSkillWork
ChatGPTRegister the user-requested LP or ad-copy task. Save only relevant user-provided premises, not whole conversations. No server generation.
stopFreeSkillImprovement
ChatGPTStop automatic review preparation only. Measurement permission and saved data are separate.
stopFreeSkillMeasurement
ChatGPTStop this explicit measurement permission; reconnecting never restarts it. Retained observations remain readable.
verifyFreeSkillDeployment
ChatGPTRead the explicitly allowlisted pilot LP and compare its HTML with the saved source. Confirmation is point-in-time, not historical coverage or external asset verification.