Customgpt Ai Ragupdate Agent

Walk up from $PWD toward /, looking for .customgpt-meta.json. Extract agent_id, agent_name, indexed_folder, and included_paths. Store the full path as $META_FILE_PATH.

If included_paths is missing, default to ["."].

If not found:

"No agent found in this directory tree. Run /create-agent to set one up first."

Step 3 — Collect All Eligible Files on Disk

Important: find is recursive — it automatically includes all subfolders. Do NOT manually iterate through subdirectories.

For each path in included_paths (resolved relative to indexed_folder):

find "${indexed_folder}/${path}" -type f \
  -not -path "*/.git/*" \
  -not -path "*/node_modules/*" \
  -not -path "*/__pycache__/*" \
  -not -path "*/.next/*" \
  -not -path "*/dist/*" \
  -not -path "*/build/*" \
  -not -path "*/.cache/*" \
  -not -path "*/vendor/*" \
  -not -path "*/coverage/*" \
  -not -path "*/.venv/*" \
  -not -path "*/venv/*" \
  -not -path "*/target/*" \
  -not -path "*/.turbo/*" \
  -not -path "*/.parcel-cache/*" \
  -not -name ".env" \
  -not -name ".env.*"

CustomGPT.ai supports 1,400+ file types — do NOT filter by extension. If a file type is unsupported, the API will return an error for that file.

If included_paths is ["."], check indexed_folder itself.

Compute REL for each file (path relative to indexed_folder). Store as the full local file list.

Step 4 — Fetch Document List from Agent

Fetch all documents currently in the agent:

curl -s --request GET \
  --url "https://app.customgpt.ai/api/v1/projects/${AGENT_ID}/pages?page=${PAGE}&limit=100&order=asc" \
  --header "Authorization: Bearer ${API_KEY}" \
  --header "accept: application/json"

Paginate through all pages. Collect all filename values into a set of known documents.

Step 5 — Categorize Files

Compare local files against the agent's document list:

• Changed files: exist in agent AND are newer than $META_FILE_PATH (use find -newer or compare timestamps)
• New files: exist on disk within included_paths but their REL path has no match in the agent's document list

For changed files, store the mapping: REL → page_id.

Step 6 — Confirm with User

If no changed files AND no new files:

"Everything is up to date — no changes detected."

Otherwise, present what was found:

"Found:"

• {CHANGED_COUNT} changed file(s) — will be re-uploaded (old version deleted first) {list of changed filenames}

If new files were found, also show:

• {NEW_COUNT} new file(s) not yet in the agent: {list of new filenames} "Would you like to add these new files as well? (yes/no/skip)"

"Changed and new files will count against your monthly processing limits. Continue? (yes/no)"

If not confirmed, stop.

Step 7 — Delete Old Versions of Changed Files

For each changed file that has an existing document ID:

curl -s --request DELETE \
  --url "https://app.customgpt.ai/api/v1/projects/${AGENT_ID}/pages/${PAGE_ID}" \
  --header "Authorization: Bearer ${API_KEY}" \
  --header "accept: application/json"

HTTP 200 = deleted. HTTP 404 = already gone (fine, continue).

Step 8 — Upload Changed + New Files

Upload all changed files (new versions) and, if the user confirmed, new files too:

curl -s --request POST \
  --url "https://app.customgpt.ai/api/v1/projects/${AGENT_ID}/sources" \
  --header "Authorization: Bearer ${API_KEY}" \
  --form "file=@${ABSOLUTE_PATH};filename=${REL}"

HTTP 200 or 201 = success. Report each result: ✓ {REL} or ✗ {REL} (HTTP {status}).

Step 9 — Reset Freshness Timestamp

touch "${META_FILE_PATH}"

Step 10 — Report

Agent '{agent_name}' updated.

• Changed files re-uploaded: {CHANGED_UPLOADED}
• New files added: {NEW_UPLOADED}
• Failed: {FAILED}{FAILED > 0 ? " — run /update-agent again to retry" : ""}

CustomGPT.ai is now processing the updated files. Run /check-status to monitor progress, then /ask-agent to search.