fix: fix: disinto init re-run silently drops HUMAN_TOKEN when token already exists (#275) #296

Merged

dev-bot merged 1 commit from fix/issue-275 into main

2026-04-06 12:24:03 +00:00

dev-bot commented

2026-04-06 12:14:57 +00:00

Collaborator

Fixes #275

Changes

Fixes #275 ## Changes

dev-bot added 1 commit 2026-04-06 12:14:57 +00:00

fix: fix: disinto init re-run silently drops HUMAN_TOKEN when token already exists (#275 )

ci/woodpecker/push/ci Pipeline was successful

Details

ci/woodpecker/pr/ci Pipeline was successful

Details

ci/woodpecker/pr/smoke-init Pipeline was successful

Details

7977e2562c

Apply delete-then-recreate pattern for human token (matching admin token in PR #274).
Forge/Forgejo only returns sha1 at creation time; listing returns no sha1, causing
HUMAN_TOKEN to be silently empty on re-runs when token name already exists.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

dev-bot commented

2026-04-06 12:22:03 +00:00

Author

Collaborator

AI Review — Error\n\nReview failed.\n---\n`7977e25`

## AI Review — Error\n\nReview failed.\n---\n*7977e25*

review-bot commented

2026-04-06 12:23:03 +00:00

Collaborator

AI Review

Summary

This PR fixes a silent data-loss bug: Forgejo only returns the token secret (sha1) at creation time, never when listing. The old fallback attempted listing tokens and reading sha1 from the result, which always yields null, causing HUMAN_TOKEN to be silently set to empty on every re-run.

Approach

Delete the existing named token first (by ID), then create a fresh one — guaranteeing the secret is always captured.

Findings

No issues found. Specific checks:

Logic: delete-then-create is the correct pattern for Gitea-family APIs where secrets are not re-retrievable. Root cause correctly identified in the comment.
Error handling: || existing_human_token_id="" covers curl/jq failures; || true on DELETE correctly tolerates 404 or transient errors without aborting.
Idempotency: First run (no existing token) skips DELETE and creates normally. Re-run deletes stale token and creates fresh. Both paths store the result in .env.
Token name filter: select(.name == "disinto-human-token") is precise — avoids accidentally deleting unrelated tokens. Forgejo enforces unique token names per user, so at most one ID is returned.
Security: No secrets echoed or logged; credentials passed via -u flag. No vault-bypass patterns detected.
Bash: Proper local declarations, quoting, and consistency with surrounding code style.

Verdict

APPROVE — Correctly fixes token-drop by delete-then-recreate instead of broken list-fallback

Reviewed at 7977e25 | AGENTS.md

## AI Review  ### Summary This PR fixes a silent data-loss bug: Forgejo only returns the token secret (sha1) at creation time, never when listing. The old fallback attempted listing tokens and reading sha1 from the result, which always yields null, causing HUMAN_TOKEN to be silently set to empty on every re-run. ### Approach Delete the existing named token first (by ID), then create a fresh one — guaranteeing the secret is always captured. ### Findings No issues found. Specific checks: - **Logic**: delete-then-create is the correct pattern for Gitea-family APIs where secrets are not re-retrievable. Root cause correctly identified in the comment. - **Error handling**: || existing_human_token_id="" covers curl/jq failures; || true on DELETE correctly tolerates 404 or transient errors without aborting. - **Idempotency**: First run (no existing token) skips DELETE and creates normally. Re-run deletes stale token and creates fresh. Both paths store the result in .env. - **Token name filter**: select(.name == "disinto-human-token") is precise — avoids accidentally deleting unrelated tokens. Forgejo enforces unique token names per user, so at most one ID is returned. - **Security**: No secrets echoed or logged; credentials passed via -u flag. No vault-bypass patterns detected. - **Bash**: Proper local declarations, quoting, and consistency with surrounding code style. ### Verdict **APPROVE** — Correctly fixes token-drop by delete-then-recreate instead of broken list-fallback --- *Reviewed at `7977e25` | [AGENTS.md](AGENTS.md)*

review-bot approved these changes 2026-04-06 12:23:03 +00:00

review-bot left a comment

Collaborator

AI Review: APPROVE — Correctly fixes token-drop by delete-then-recreate instead of broken list-fallback

AI Review: **APPROVE** — Correctly fixes token-drop by delete-then-recreate instead of broken list-fallback