fix(strands-command): return full PR diffs and offload large tool results#63
Open
yonib05 wants to merge 6 commits into
Open
fix(strands-command): return full PR diffs and offload large tool results#63yonib05 wants to merge 6 commits into
yonib05 wants to merge 6 commits into
Conversation
get_pr_files capped every file's patch at 50 lines, so the reviewer silently lost the tail of any larger change and reasoned about code it never saw. Replace the per-file line cap with a total character budget: inline full patches until the budget is reached, then list remaining files for on-demand fetch instead of truncating mid-file.
…ping it A file whose patch alone exceeds the budget previously produced no inline diff at all, leaving the reviewer dependent on fetching it. Inline a line-trimmed head slice when enough budget remains so there is always some inline signal, and note the omitted remainder is fetchable. Adds tests for the partial-head and mixed inline/overflow cases.
Guard the partial-head branch so a patch starting with a newline (whose line-boundary trim leaves an empty head) is fully deferred instead of emitting an empty "Diff (head only):" block. Add tests covering the full-deferral branch, the exact budget boundary, and the empty-head edge, merge the two near-duplicate over-budget tests, and note the accepted file-order dependence of the diff budget.
This was referenced Jun 15, 2026
zastrowm
reviewed
Jun 19, 2026
| if head: | ||
| used += len(head) | ||
| omitted = len(patch) - len(head) | ||
| overflow.append(f"{filename} (+{additions} -{deletions}, partially shown)") |
Member
There was a problem hiding this comment.
Can we write the overflow to disk and let the agent know where it is?
|
|
||
| @tool | ||
| @log_inputs | ||
| def get_pr_files(pr_number: int, repo: str | None = None) -> str: |
Member
There was a problem hiding this comment.
This is PR output?
Can we just leverage the Strands Tool Offloader here and let it manage the truncation?
Member
Author
There was a problem hiding this comment.
Good call will swap it should simplify alot
Member
Author
There was a problem hiding this comment.
Done in 52c3ecf. Swapped the bespoke budgeting for the SDK's ContextOffloader plugin (wired in agent_runner.py with InMemoryStorage). get_pr_files now just returns the full diff, and the plugin offloads any oversized tool result — not only diffs but file reads and shell output too — registering a retrieval tool so the agent fetches full content on demand. Net: ~60 lines of truncation logic + its constants removed, and truncation is handled generically for every tool.
…oader Replace the bespoke per-call diff budgeting in get_pr_files with the SDK's ContextOffloader plugin. get_pr_files now returns the full untruncated diff; the plugin offloads any oversized tool result (diffs, file reads, shell output) to storage and registers a retrieval tool so the agent fetches the full content on demand. This handles truncation generically for every tool rather than just this one, per review feedback.
… sessions resolve references
yonib05
changed the title
…reviewed GitHub list endpoints cap at 100 items per page, so reading a single response silently dropped changed files past the first page on large PRs -- the reviewer would never see them. Add a paginated GET helper that follows Link rel="next" (bounded by a page cap) and use it for the PR files list.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
get_pr_fileshad two ways of hiding code from the reviewer:# Limit diff size to avoid overwhelming output), so issues below line 50 were invisible.Change
get_pr_filesnow returns each file's full, untruncated diff, and follows pagination (new_github_get_all_pageshelper, which followsLink rel="next"with a page-count safety cap) so every changed file is included.ContextOffloaderplugin (agent_runner.py). When a result exceeds the token threshold, the plugin offloads it to storage, replaces it in-context with a preview + reference, and registers aretrieve_offloaded_contenttool so the agent pulls the full content on demand.This handles oversized results generically for every tool (diffs, file reads, shell output), not just
get_pr_files, and removes ~60 lines of bespoke per-call budgeting/head-slicing logic.Storage backend
Uses
S3Storage, reusing the bucket already configured for session management. The offloaded result is replaced in the conversation by a storage reference that the session manager persists; a resumed session in a later process must still resolve that reference, which in-memory storage could not guarantee.Tests
tests/test_github_tools.py(get_pr_files):Full unit suite:
38 passed.