Configuration and Knowledge Management·Task 5.2·Bloom: analyse·Difficulty 4/5·9 min read·Updated 2026-07-14

Choosing Connector Access vs Manual Upload (CCAO-F)

Manage uploaded knowledge and connectors (e.g., Google Drive, Gmail)

SUBy Solomon UdohReviewed by Solomon UdohAI-assisted · human-reviewed
In short
Choosing between connector access and manual upload is a decision based on how frequently a source changes and who needs access. Frequently changing, authoritative source systems, such as a live Drive folder or an inbox, are better reached via a connector than through repeated manual uploads. Stable reference material that rarely changes is well suited to a one-time upload into the knowledge base. The connected set should stay deliberate, and manual uploads require an explicit re-upload step to stay current.

One decision, driven by change

Both a connector and a manual upload can put a source in front of Claude, so the practical question is which to use for a given source. The CCAO-F exam treats this as an analyse-level decision, and the dominant factor is simple: how frequently does the source change?

A connector reaches a live external system and reflects its current state. A manual upload is a static snapshot that stays as-is until someone replaces it. So the decision reduces to matching the mechanism to the source's rate of change: live reach for what moves, a fixed snapshot for what does not. Getting this right avoids both a maintenance treadmill and unnecessary connector sprawl.

Connector access versus manual upload
The decision of whether a knowledge source should be reached live through a connector or added as a static uploaded document, driven mainly by how frequently the source changes and who needs access. Frequently changing, authoritative sources suit connectors; stable reference material suits one-time uploads. Manual uploads require an explicit re-upload to stay current, while the connected set should be kept deliberate rather than maximal.

When a connector is the right call

If a source changes frequently and is authoritative, reach it with a connector. A live Drive folder that the team edits daily, or an inbox that fills continuously, is a moving target; a connector queries the current state each time, so Claude always sees what is there now. The alternative, re-uploading the document every time it changes, is a maintenance treadmill that guarantees staleness the moment someone forgets a re-upload.

The "who needs access" dimension reinforces this. When many people rely on a source that is itself the authoritative system of record, connecting to that live system keeps everyone working from the current truth rather than from someone's dated export. For frequently changing, authoritative, widely used sources, the connector is the mechanism that keeps Claude current without human effort per change.

When a one-time upload is the right call

If material is stable and rarely changes, a one-time upload into the knowledge base is the better fit. A fixed policy, a finalized reference document, a style guide that turns over once a year: uploading it once is simpler than standing up and governing a connector, and there is little maintenance cost because the source is not moving.

Here the maintenance property of uploads is an acceptable, low burden rather than a treadmill. Uploaded knowledge only changes when you re-upload it, which is exactly what its maintenance depends on, and for a rarely changing document that manual step is infrequent. This is also where curation applies: on the rare occasion the stable document is superseded, remove the old version as you add the new one so the base holds a single authoritative copy.

Choosing between a connector and a manual upload
Loading diagram...
Rate of change drives the choice: live reach for moving sources, a static snapshot for stable ones, with the connected set kept curated.

What the CCAO-F exam trips candidates on

The exam sets two traps at opposite ends. The first is manually re-uploading a frequently changing document over and over instead of connecting Claude to its live source. The scenario shows someone diligently re-uploading a daily-changing file; the credited answer replaces that treadmill with a connector to the live source. Effortful re-uploading of a moving target is the tell.

The second is connecting a broad, ungoverned set of external systems when a narrow, deliberate connector set would suffice. This is the over-reach direction: reaching for connectors everywhere, including for stable material that a single upload would handle, and granting access to everything. The credited answer keeps the connected set deliberate and curated, connecting the frequently changing sources that need it and uploading the stable ones that do not.

Worked example

A team relies on two sources: a shared Drive folder of project files that changes several times a day, and the company's data-retention policy, which is revised once a year. Someone has been re-uploading both into the Project every week. Recommend the right mechanism for each and explain the trade-off.

The two sources sit at opposite ends of the change axis, so they call for opposite mechanisms.

The shared Drive folder changes several times a day, so it is a frequently changing, authoritative source best reached with a connector. A connector queries the live folder and reflects its current state, so Claude always sees the latest files without anyone re-uploading. Re-uploading a folder that changes daily is a treadmill that is stale the moment the folder changes again, which is exactly the first trap.

The data-retention policy is revised once a year, so it is stable reference material well suited to a one-time upload into the knowledge base. Standing up a connector for a document that barely moves is unnecessary reach, and the low maintenance cost, re-uploading only when the annual revision lands, is entirely acceptable. The trade-off is change rate versus maintenance: connect what moves to avoid the treadmill, upload what is stable to avoid needless connector sprawl, and keep the connected set deliberate on both counts.

Common misreadings to avoid

Misconception

Re-uploading a document whenever it changes is a fine way to keep a fast-moving source current.

What's actually true

For a frequently changing source, repeated manual uploads are a treadmill that goes stale between re-uploads. A connector reaches the live source and stays current without per-change effort.

Misconception

If connectors keep sources current, you should connect as many sources as possible.

What's actually true

The connected set should stay deliberate and curated. Stable material is better as a one-time upload, and connecting everything creates ungoverned reach. Match the mechanism to how often each source changes.

How this shows up on the exam

Domain 5 analysis questions give you a mix of sources and ask which to connect and which to upload. Sort by rate of change: frequently changing and authoritative goes to a connector, stable reference goes to a one-time upload, and the connected set stays deliberate throughout.

This decision draws on connector capability boundaries, since a connector only fits if it can perform what the source needs, and on curating uploaded knowledge, which governs the upload path. It completes the connectors-and-knowledge picture begun in connectors as authorized external reach.

Check your understanding

A team uses a shared Drive folder that changes several times a day and a data-retention policy revised once a year. Which mechanism assignment is best?

People also ask

When should I use a connector instead of uploading a document?
When the source changes frequently and is authoritative, such as a live Drive folder or an inbox. A connector reaches the current state, so you avoid repeatedly re-uploading a changing document.
When is a one-time upload the right choice?
When the material is stable reference that rarely changes, such as a fixed policy. A one-time upload into the knowledge base suits it with little maintenance cost.
What determines connector versus upload?
Mainly how frequently the source changes, plus who needs access. Frequently changing sources favor connectors; stable references favor uploads. Keep the connected set deliberate either way.

Watch and learn

Official Anthropic Academy lessons first, then hand-picked walkthroughs. Videos load only when you press play.

No videos curated for this concept yet

We are still curating the best official and community videos for this topic.

Official prep for this domain

Anthropic's own free prep module for this part of the syllabus, on the official prep course. Free with an Anthropic Academy sign-in.

References & primary sources

Adaptive study

Master this concept with Archie

Practice it inside an adaptive study session. Archie, your Socratic AI tutor, tracks your mastery with Bayesian Knowledge Tracing and schedules the perfect next review.

Start studying