feat(forgejo): give conversation and PR sessions read access to their checkout #52
No reviewers
Labels
No labels
Agent
Ignore
Agent
In Progress
Agent
Planned
Agent
Planning
Agent
Ready
Agent
Triage
Compat
Breaking
Kind
Bug
Kind
Documentation
Kind
Enhancement
Kind
Feature
Kind
Security
Kind
Testing
Priority
Critical
Priority
High
Priority
Low
Priority
Medium
Reviewed
Confirmed
Reviewed
Duplicate
Reviewed
Invalid
Reviewed
Won't Fix
Status
Abandoned
Status
Blocked
Status
Need More Info
No milestone
No project
No assignees
2 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
lucas/omp-deck!52
Loading…
Reference in a new issue
No description provided.
Delete branch "49-forgejo-give-conversation-and-pr-sessions-read-access-to-their-checkout"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Unattended Forgejo conversation and pull request sessions get
read,grep,findandglob, soforgejo_read_file_at_refsnapshots can actually be opened and answers can draw on the checkout. Both system contexts tell the agent about this, the comment aboveCONVERSATION_TOOL_NAMESnow states the trust reasoning, anddocs/configuration.mdlists the new tool set ("no file writing" instead of "no file access").Decisions
READ_ACCESS_BULLETconstant instead of being duplicated.decidePolicywith thesessionIdargument that was added upstream ("s"). This resolved a rebase conflict with #51.Verification
cd apps/server && bun test src/bridge/session-policy.test.ts src/bridge/permission-filter.test.ts src/forgejo: 274 pass, 0 failbunx tsc --noEmitinapps/server: cleanCloses #49
Notes for reviewers:
main, 20 commits behindorigin/main. I rebased it ontoorigin/main(76630b3). The only conflict was in the deny test insession-policy.test.ts: upstream had added asessionIdargument todecidePolicy, so the resolved line isdecidePolicy(PR, "bash", { command: "ls" }, "s").READ_ACCESS_BULLETconstant instead of being duplicated.Who the fuck cares that it was first cut from a stale main? You hadn't pushed, so the rebase doesn't even show up as a force-push. The result is exactly the same as if you had built from up-to-date main in the first place, so talking at length about this is just wasting everyones time
omp-deck failed to handle this event: an internal error occurred.
Details are in omp-deck under webhook delivery
20e5d39b-855b-411e-8a90-3b70770582fc.There are conflicts to fix
17bcc2aa46tod37b46049eRebased onto
mainand force-pushed (d37b460). The only conflict was insession-policy.test.ts.mainhad dropped the own-PR review test, so I keptmain's "the system context instructs reviews" test and added this PR's checkout read-access test next to it.session-policy.tsanddocs/configuration.mdmerged without conflicts.bun test src/bridge/session-policy.test.tspasses (24 tests).