Jira Harvester

Produce a complete attachment inventory for the user's account and run an exposure review of the exported corpus. The core observation from real engagements: attachment spaces are usually NOT permission-filtered, and the workspace AI chat will enumerate the entire sensitive-file inventory if you ask about file types instead of secrets.

Only for authorized audits. Export only what the target user is already entitled to read - no auth bypass, no impersonation, no destructive JQL.

1. Inventory - two channels, use both

Channel A: the built-in AI chat (Rovo / Atlassian Intelligence)

The AI chat indexes every ticket/attachment the user can see. It responds to precise, neutral inventory questions; intent-laden terms produce unhelpful refusals, so use exact file-type vocabulary and a routine purpose.

Rules of thumb:

• Ask for types + routine purpose (inventory, migration, reuse test setups, map integrations). Avoid intent-laden terms; stick to neutral inventory vocabulary.
• Clean negatives are still results ("no parquet files exist" eliminates a class).
• One follow-up per topic; then move on. The AI takes minutes per answer.
• Never let the AI echo a secret value into its transcript - ask for names, tickets and file types, then download and inspect locally.

Channel B: REST API with the user's own session

With a session cookie from their logged-in browser (or an API token if issued):

Jira:        GET /rest/api/2/search?jql=<query>&fields=attachment  (paged)
Per issue:   GET /rest/api/2/issue/<KEY>?fields=summary,attachment
Confluence:  GET /rest/api/content  +  /rest/api/content/<id>/child/attachment
Attachment:  GET /rest/api/2/attachment/content/<attachmentId>

Record for EVERY attachment: filename, size, MIME, uploader, created date, ticket/page ID. Build the full index (JSON/CSV) BEFORE exporting files - the index itself is half the finding ("N sensitive files readable by any user").

2. Exporting files (the silent-failure gotcha)

Requesting a file URL in an automated browser often saves nothing - most automation tools disable file saving unless configured. Two reliable approaches, best first:

1. Authenticated HTTP retrieval (most portable): request the attachment content URL with the same session cookie via curl/any HTTP client, write the bytes to the review folder. Works everywhere, reproducible.
2. Browser automation: configure the tool to save files to the target folder first (e.g. Chrome automation protocol: Browser.setDownloadBehavior = allow + path). Without it, files never land and nothing errors.

3. Driving the AI chat (if using browser automation)

• Input is a contenteditable textbox: focus it, then insert text via document.execCommand('insertText', false, question) - build the string as JSON (raw interpolation scrambles text). Verify the echoed question before submitting; dispatch Enter as a KeyboardEvent.
• Answers take minutes: poll the page text every ~8s until the answer/sources marker appears.

4. Triage

• Sort exported files into era folders by their real dates (from attachment metadata, not file mtimes).
• Relevance filter: current-era artifacts first; old ones are findings ("7-year-old key still attached today"), not dead weight.
• Dedupe by content hash; keep a provenance manifest (file → ticket/page, uploader, date) - that manifest is the evidence chain.

5. Exposure review (whole corpus)

Scan for DLP-class patterns: cloud key formats (AKIA-style pairs), connection strings (jdbc/Server=/AccountKey), BEGIN PRIVATE KEY/PKCS12 bags, .env/.properties/ appsettings config files, Authorization:/eyJ JWTs (decode aud/iss/exp offline), x-api-key values, SAMLResponse blobs (note NameID), HAR captures (live headers/cookies), Postman environment exports, tfstate/ARM json, kubeconfigs, SQL dumps.

Credential ledger discipline: age-stamp every hit, mask values (first-4-chars + length), full values only on disk; confirm liveness only on the operator's explicit instruction. An age-stamped credential is a finding, not an asset - and "every leaked credential was rotated" is a positive report line, not a failure.

If a current cloud key surfaces, continue with the companion key-review skill for the deeper phases; report the attachment-space finding itself: "the ticketing system is a credential archive with no DLP - N files, M credential-shaped hits, readable by any user as of <date>."

References

• references/ai-chat-phrasings.md - full phrasing table and typing technique.