feat(enable): add --limit flag to cap issues imported from beads/GitHub#252
feat(enable): add --limit flag to cap issues imported from beads/GitHub#252visigoth wants to merge 1 commit intofrankbria:mainfrom
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (4)
🚧 Files skipped from review as they are similar to previous changes (1)
WalkthroughAdded an optional per-source Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant User as "User (CLI)"
participant Script as "ralph_enable / ralph_enable_ci"
participant BeadsCLI as "bd CLI"
participant GitHubCLI as "gh CLI / API"
participant Parser as "jq / local parser"
User->>Script: run import with --limit N
Script->>BeadsCLI: bd list --limit N
BeadsCLI->>Parser: emit JSON / lines
Parser->>Script: parsed Beads tasks
alt limit == "0"
Script->>GitHubCLI: gh api --paginate /repos/.../issues?state=open&per_page=100
GitHubCLI->>Parser: emit pages
Parser->>Script: filter out items with "pull_request"
else limit > 0
Script->>GitHubCLI: gh issue list --limit N --json ...
GitHubCLI->>Parser: emit JSON
end
Parser->>Script: parsed GitHub tasks
Script->>Script: merge/import tasks into pipeline
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
lib/task_sources.sh (1)
173-175:⚠️ Potential issue | 🟡 MinorUpdate stale parameter docs for
fetch_github_tasks.Line 174 still says default is
50, but Line 186 now defaults to0. Please align the function comment with implementation.As per coding guidelines "Update inline comments in bash scripts immediately when implementation changes, remove outdated comments".
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@lib/task_sources.sh` around lines 173 - 175, The parameter documentation for fetch_github_tasks is out of sync: the comment for $2 (limit) says default 50 but the implementation defaults to 0; update the inline comment block above fetch_github_tasks so the $2 (limit) description matches the actual default used in the function (change "default: 50" to "default: 0" or adjust the function to use 50 if that was intended), and ensure any surrounding explanatory text is consistent with the implementation.
🧹 Nitpick comments (1)
ralph_enable_ci.sh (1)
338-347: Declare task buffers aslocalinmain()to avoid global leakage.Line 338 and Line 345 assign
beads_tasks/github_taskswithout local declaration, which can leak state in longer scripts.Scoped variable cleanup
- local imported_tasks="" + local imported_tasks="" + local beads_tasks="" + local github_tasks="" + local prd_tasks="" case "$TASK_SOURCE" in beads) if beads_tasks=$(fetch_beads_tasks "open" "$IMPORT_LIMIT" 2>/dev/null); thenBased on learnings "Bash script naming conventions: use snake_case for function names and local variables ... ensure local variables inside functions are declared with the local keyword".
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@ralph_enable_ci.sh` around lines 338 - 347, The temporary task buffers beads_tasks and github_tasks are being assigned inside main() without local, leaking into global scope; change their assignments to declare them as local (e.g., use local beads_tasks=$(fetch_beads_tasks "open" "$IMPORT_LIMIT" 2>/dev/null) and local github_tasks=$(fetch_github_tasks "$GITHUB_LABEL" "$IMPORT_LIMIT" 2>/dev/null)) or add local declarations immediately before their current assignments so the variables are function-scoped and do not persist after main() returns.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@lib/task_sources.sh`:
- Around line 186-189: The current code remaps limit=0 to 1000 which silently
truncates imports; update the function to honor "0 = all" by removing the
hard-cap mapping of the local variable limit and instead fetch issues with
automatic pagination using gh api /repos/OWNER/REPO/issues?state=open --paginate
(or equivalent gh api call) when limit is 0, and use gh issue list --limit when
a positive numeric limit is supplied; also adjust the help text that currently
says "default: 50" to reflect the real default "0" (all) so the CLI help and the
variable behavior match.
---
Outside diff comments:
In `@lib/task_sources.sh`:
- Around line 173-175: The parameter documentation for fetch_github_tasks is out
of sync: the comment for $2 (limit) says default 50 but the implementation
defaults to 0; update the inline comment block above fetch_github_tasks so the
$2 (limit) description matches the actual default used in the function (change
"default: 50" to "default: 0" or adjust the function to use 50 if that was
intended), and ensure any surrounding explanatory text is consistent with the
implementation.
---
Nitpick comments:
In `@ralph_enable_ci.sh`:
- Around line 338-347: The temporary task buffers beads_tasks and github_tasks
are being assigned inside main() without local, leaking into global scope;
change their assignments to declare them as local (e.g., use local
beads_tasks=$(fetch_beads_tasks "open" "$IMPORT_LIMIT" 2>/dev/null) and local
github_tasks=$(fetch_github_tasks "$GITHUB_LABEL" "$IMPORT_LIMIT" 2>/dev/null))
or add local declarations immediately before their current assignments so the
variables are function-scoped and do not persist after main() returns.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 98423a79-cd8f-479f-a7e6-041a6acd8a34
📒 Files selected for processing (3)
lib/task_sources.shralph_enable.shralph_enable_ci.sh
visigoth
force-pushed
the
fix/beads-limit
branch
from
65c3299 to
2cce79c
Compare
There was a problem hiding this comment.
Actionable comments posted: 1
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
ralph_enable_ci.sh (1)
339-359:⚠️ Potential issue | 🟠 MajorFix zero-task counting before emitting CI output.
On Lines 343, 350, and 358,
grep -calready prints0when nothing matches, so|| echo "0"turnsTASKS_IMPORTEDinto0\n0. With--limitor a label filter that yields no tasks, the CI summary and--jsonoutput become malformed.Suggested fix
- TASKS_IMPORTED=$(echo "$imported_tasks" | grep -c '^\- \[' || echo "0") + TASKS_IMPORTED=$(printf '%s\n' "$imported_tasks" | awk '/^- \[/{count++} END{print count+0}') @@ - TASKS_IMPORTED=$(echo "$imported_tasks" | grep -c '^\- \[' || echo "0") + TASKS_IMPORTED=$(printf '%s\n' "$imported_tasks" | awk '/^- \[/{count++} END{print count+0}') @@ - TASKS_IMPORTED=$(echo "$imported_tasks" | grep -c '^\- \[' || echo "0") + TASKS_IMPORTED=$(printf '%s\n' "$imported_tasks" | awk '/^- \[/{count++} END{print count+0}')🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@ralph_enable_ci.sh` around lines 339 - 359, The TASKS_IMPORTED assignment in the TASK_SOURCE case blocks uses "grep -c ... || echo '0'", causing duplicate "0\n0" when grep -c already returns 0; remove the "|| echo '0'" fallback so TASKS_IMPORTED is set directly from the grep output. Update the three locations where TASKS_IMPORTED is assigned (the beads, github, and prd branches that use fetch_beads_tasks, fetch_github_tasks, and extract_prd_tasks) to use TASKS_IMPORTED=$(echo "$imported_tasks" | grep -c '^\- \[') and leave the rest of the output_message calls unchanged.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@lib/task_sources.sh`:
- Around line 221-235: The code can silently truncate results when a label
filter is used because `gh issue list` switches to the Search API which is
capped at 1000; update the logic around the gh_args construction (the label
variable and limit variable usage) to validate that when label is non-empty the
requested limit is <=1000 and fail fast with a clear error (return non-zero and
message) if it exceeds that, or alternatively replace the bounded `gh issue
list` branch with a paginated `gh api graphql` cursor-based fetch to honor
larger limits; implement the simpler fix by adding a check after the `if [[ -n
"$label" ]]; then` block that inspects `limit` and returns 1 with an explanatory
message if limit > 1000.
---
Outside diff comments:
In `@ralph_enable_ci.sh`:
- Around line 339-359: The TASKS_IMPORTED assignment in the TASK_SOURCE case
blocks uses "grep -c ... || echo '0'", causing duplicate "0\n0" when grep -c
already returns 0; remove the "|| echo '0'" fallback so TASKS_IMPORTED is set
directly from the grep output. Update the three locations where TASKS_IMPORTED
is assigned (the beads, github, and prd branches that use fetch_beads_tasks,
fetch_github_tasks, and extract_prd_tasks) to use TASKS_IMPORTED=$(echo
"$imported_tasks" | grep -c '^\- \[') and leave the rest of the output_message
calls unchanged.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 6ac8a1e6-71a6-4fd4-bc9b-4172574df143
📒 Files selected for processing (3)
lib/task_sources.shralph_enable.shralph_enable_ci.sh
| else | ||
| # Bounded fetch via `gh issue list` which already excludes PRs | ||
| local gh_args=("issue" "list" "--state" "open" "--limit" "$limit" "--json" "number,title,labels") | ||
| if [[ -n "$label" ]]; then | ||
| gh_args+=("--label" "$label") | ||
| fi | ||
| if ! json_output=$(gh "${gh_args[@]}" 2>/dev/null); then | ||
| return 1 | ||
| fi | ||
| if command -v jq &>/dev/null; then | ||
| tasks=$(echo "$json_output" | jq -r ' | ||
| .[] | | ||
| "- [ ] [#\(.number)] \(.title)" | ||
| ' 2>/dev/null) | ||
| fi |
There was a problem hiding this comment.
🧩 Analysis chain
🌐 Web query:
In GitHub CLI, is gh issue list --limit Nhard-capped at 1000 even whenN > 1000? If so, what is the recommended way to fetch exactly the first N open issues when N may exceed 1000?
💡 Result:
No, gh issue list --limit N is not hard-capped at 1000 when N > 1000 for basic usage without filters that trigger the Search API (like --label or --search). It uses the GitHub REST API's List issues endpoint, which supports pagination up to the specified limit via GraphQL or repeated REST calls. However, certain filters switch to the Search API, which caps at 1000 results total. The recommended way to fetch exactly the first N open issues (sorted by most recent first, default) when N may exceed 1000 is to use gh issue list --state open --limit N. For repos with more than 1000 open issues, set N large enough to cover all (e.g., estimate or use a very high number like 10000), as there is no built-in unbounded fetch; users must specify a limit intentionally. To fetch ALL open issues reliably, use gh api graphql --paginate with a query like: query($owner: String!, $repo: String!, $cursor: String) { repository(owner: $owner, name: $repo) { issues(first: 100, after: $cursor, states: OPEN, orderBy: {field: CREATED_AT, direction: DESC}) { pageInfo { hasNextPage endCursor } nodes { number title } } } } This handles cursor pagination without limits other than API constraints.
Citations:
- 1: Update documentation around
--limitflag restriction cli/cli#4888 - 2: https://cli.github.com/manual/gh_issue_list
- 3: gh pr list limitation ? cli/cli#3836
- 4: bypass --limit to fetch all issues cli/cli#9522
Guard against silent result truncation when --label filtering is combined with high limits.
When --label is present in the query (lines 224-226), gh issue list switches to the Search API, which is hard-capped at 1000 results regardless of the --limit value. A call requesting --limit 5000 with a label filter will silently return only 1000 issues. Either validate that limit <= 1000 when a label filter is active, or use paginated gh api graphql with cursor pagination to fetch beyond 1000 results when needed.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@lib/task_sources.sh` around lines 221 - 235, The code can silently truncate
results when a label filter is used because `gh issue list` switches to the
Search API which is capped at 1000; update the logic around the gh_args
construction (the label variable and limit variable usage) to validate that when
label is non-empty the requested limit is <=1000 and fail fast with a clear
error (return non-zero and message) if it exceeds that, or alternatively replace
the bounded `gh issue list` branch with a paginated `gh api graphql`
cursor-based fetch to honor larger limits; implement the simpler fix by adding a
check after the `if [[ -n "$label" ]]; then` block that inspects `limit` and
returns 1 with an explanatory message if limit > 1000.
|
All three CodeRabbit findings addressed in the force-pushed commit:
Full test suite passes (the two failing tests — BSD stat fallback in |
Adds a --limit <n> flag to ralph-enable and ralph-enable-ci that controls the maximum number of issues fetched from any source. Default is 0 (all issues), which fixes the previous behavior where bd list silently capped at 50 items and gh issue list at 50. The limit is threaded through fetch_beads_tasks, fetch_github_tasks, get_beads_count, and import_tasks_from_sources. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
visigoth
force-pushed
the
fix/beads-limit
branch
from
2cce79c to
9775288
Compare
|
Also added 4 regression tests for
Tests use a mock |
frankbria
left a comment
frankbria
left a comment
There was a problem hiding this comment.
The --limit feature is a good addition. A few issues to fix before it can merge:
Blocking
Same grep -c || echo "0" bug in newly added code
The PR being fixed by #251 is grep -c ... || echo "0" producing "0\n0". The new code in ralph_enable_ci.sh (~lines 343, 350, 358) introduces the same pattern for TASKS_IMPORTED. Since #251 fixes the pattern elsewhere, this branch needs to use the safe form too:
# Instead of:
TASKS_IMPORTED=$(echo "$imported_tasks" | grep -c '^\- \[' || echo "0")
# Use:
TASKS_IMPORTED=$(echo "$imported_tasks" | grep -c '^\- \[') || TASKS_IMPORTED=0gh issue list silently truncates when using label filters
When limit=0 ("import all") is combined with a label filter, the code maps 0 → 1000 and uses gh issue list --limit 1000. GitHub's label-filtered list endpoint caps at 1000, so repos with more than 1000 matching issues will be silently truncated. Either document this cap clearly, or use paginated gh api calls when limit=0.
Minor (please fix)
Stale comment in fetch_github_tasks
Line ~174 in lib/task_sources.sh still says the $2 (limit) default is 50, but the implementation defaults to 0. Update the comment to match.
local declarations missing in main()
beads_tasks and github_tasks are assigned in ralph_enable_ci.sh:main() without local, leaking into global scope. Add local beads_tasks="" / local github_tasks="" before their first use.
Also: please rebase
PR #256 (a 1-line crash fix in ralph_loop.sh) is being merged first. Please rebase this branch onto main after that lands.
Once these are addressed, happy to re-review.
2 participants
Summary
Adds a
--limit <n>flag toralph-enableandralph-enable-cithat bounds the number of issues imported from each task source (beads, GitHub) during project enablement. Default is0, meaning "import all" — preserving today's behavior.Motivation: on repos with large beads or GitHub issue backlogs, the enable wizard currently dumps every open issue into
fix_plan.md, producing an unwieldy initial task list.--limit 25lets the user bring in just the top of the queue and grow from there.Changes
lib/task_sources.shfetch_beads_tasksnow takes a secondlimitargument and forwards--limittobd list(both JSON and text fallback paths).get_beads_countaccepts alimitargument symmetrically.fetch_github_taskshonors thelimitargument, mapping0→1000sincegh issue listtreats0differently from "all".import_tasks_from_sourcesplumbs the limit through to both source fetchers.ralph_enable.sh— adds--limit <n>CLI parsing with non-negative-integer validation, default0, plumbed intofetch_beads_tasks/fetch_github_tasks.ralph_enable_ci.sh— same--limit <n>flag for the non-interactive path, withoutput_errorvalidation.Why it's safe
0is a no-op for existing users.Test plan
ralph-enable --from beads --limit 5imports at most 5 beads issues intofix_plan.mdralph-enable --from github --limit 10imports at most 10 GitHub issuesralph-enable --from beads(no--limit) still imports the full backlogralph-enable --limit -3andralph-enable --limit fooexit with the validation errorralph-enable-ci --from beads --limit 5 --jsonproduces the limited list and a clean JSON envelopenpm testpasses locallySummary by CodeRabbit