Find what bears on the decision
Use indexed material to investigate a question across documents and discussions. Bring relevant passages and source references into the developing skill.
Explore the research workflowBring documents, recordings, transcripts, and diagrams into AI Agent Skill development. Extract usable text, distinguish speakers, and organize source material for research and review.
Make the source part of the reasoning.
01 / Start with the source
An expert’s explanation often points to a release note, a recording, a diagram, or an earlier piece of work. Bring those sources into the project so the skill can be developed with the evidence available for examination.
Knowledge Source Intake brings the material together, prepares usable text, and makes it searchable for research. The question guiding the skill gives that collection a purpose.
Start with the skill’s evidence needs02 / Explore the material
Bring PDFs, documentation, audio, video, transcripts, and supporting diagrams into the same skill project. Choose a source below to see what it contributes and which question it leaves for review.
“This release removes the legacy account status field.”
Look for the release identifier, the change description, and any recovery constraints.
The notes identify a change the expert’s answer needs to account for. Searchable text helps researchers find the relevant statement and compare it with other source material.
Does the previous application version depend on the field that was removed?
“We held the rollback because the earlier version still expected that field.”
Keep the statement connected to the speaker’s turn and the surrounding discussion.
A transcript makes the discussion available for research. Distinguishing speaker turns helps preserve the difference between a question, an expert’s explanation, and a response from someone else.
What evidence did the speaker use to establish the compatibility issue?
Keep the diagram alongside the expert’s explanation of the relationships it shows.
The visual source provides system context for the explanation. Use it with the expert’s written reasoning to identify which dependencies could change the recovery decision.
Does this map reflect the affected release, and which dependency matters to the proposed recovery?
03 / Read the discussion
Expertise often arrives in a conversation. When a transcript is missing, prepare one from the recording, distinguish speakers, and separate the discussion into evidence chunks for the skill project.
Use extracted text to find an explanation, a constraint, or an unresolved question in an audio or video recording.
Preserve who was asking, explaining, or responding. Confirm speaker attribution and transcript wording during review.
A short statement may depend on the question before it. Carry that context into research and reasoning analysis.
“What stopped us from rolling back?”
“The earlier version still expected the removed field.”
“Then the recovery path needs a compatibility check.”
04 / Prepare for review
A searchable source gives researchers somewhere to look. Skill development continues by examining the explanation, checking the evidence, and pursuing the context the source leaves unresolved.
Use indexed material to investigate a question across documents and discussions. Bring relevant passages and source references into the developing skill.
Explore the research workflowAnalyze the captured explanation for decision rationales, constraints, and exceptions. Keep the source context available as the reasoning is developed.
Explore Reasoning AnalysisConnect findings to their sources, examine dependencies, and resolve contradictions. Determine what is sufficient and what still needs an expert’s input.
Explore AI Skill ValidationStart with the evidence you have
Bring a workflow and the documents, recordings, or diagrams your experts rely on. We can help connect that material to a focused AI Agent Skill development project.