[docs] Update documentation for features from 2026-05-10

[github-actions]

Documentation Updates - 2026-05-10

This PR updates the documentation based on user-facing features merged in the last 24 hours.

Features Documented

• printf added to the default bash allow-list (from Add printf to default bash allow-list to unblock safeoutputs noop pipelines #31286)
• gemini-flash-lite model alias added; mini meta-alias re-pointed to it (from [model-inventory] Add gemini-flash-lite alias and register 2026-05-10 multiplier gaps #31283)

📝 Detailed Changes & References

Changes Made

• docs/src/content/docs/reference/tools.md — updated the documented Bash tool defaults from (echo, ls, pwd, cat, head, tail, grep, wc, sort, uniq, date) to (echo, printf, ls, pwd, cat, head, tail, grep, wc, sort, uniq, date, yq) to match pkg/constants/tool_constants.go::DefaultBashTools. printf was the addition from Add printf to default bash allow-list to unblock safeoutputs noop pipelines #31286; yq was already present in source but had been omitted from the docs, so it is included here for accuracy.
• docs/src/content/docs/reference/model-alias-specification.md — added a gemini-flash-lite row in §9.1 Vendor Family Aliases and updated the mini meta-alias in §9.2 to resolve to gemini-flash-lite instead of gemini-flash, matching pkg/workflow/data/model_aliases.json after [model-inventory] Add gemini-flash-lite alias and register 2026-05-10 multiplier gaps #31283. The auto-generated docs/src/content/docs/reference/model-tables.md already reflected these values; this PR brings the hand-written specification document into sync.

Merged PRs Referenced

• Add printf to default bash allow-list to unblock safeoutputs noop pipelines #31286 — Add printf to default bash allow-list to unblock safeoutputs noop pipelines
• [model-inventory] Add gemini-flash-lite alias and register 2026-05-10 multiplier gaps #31283 — Add gemini-flash-lite alias and register 2026-05-10 multiplier gaps

Skipped Issues

N/A — gh issue list --label documentation calls returned 403 from the API during this run, so no open or recently closed documentation issues could be enumerated. If any such issues exist, they were not considered in this run and should be triaged in a follow-up.

Notes

• Several other PRs merged in the last 24 hours (Fix auto-triage noop compliance and unbloat-docs pre-flight skip logic #31266, Refactor main_entry integration tests to idiomatic testify assertions #31259, fix(schema-consistency-checker): replace max-turns with max-effective-tokens 20M and reduce turn waste #31258, Refactor audit_agent_output_test.go to idiomatic testify assertions #31255, fix(static-analysis): stop recreating closed RGS-* issues daily — dedup by rule+file across open and closed states #31254, Constrain Copilot PR NLP workflow to never install Python deps during agent turns #31253, Prevent placeholder discussion posts in daily compiler quality reports #31252, Fix Daily SPDD cache-memory persistence guidance to avoid denied shell writes #31247, Prevent duplicate weekly discussions in Agent Performance Analyzer #31246, Codex startup: fail fast on missing Node runtime; provision Node in daily-fact workflow #31245, fix: add rate-limit retry to PR creation and fallback issue paths #31244, Harden GH_AW_MCP_CLI_SERVERS shell export to resolve code scanning alert #580 #31238, Harden release workflow with retry/backoff for release ID resolution #31236, [caveman] Optimize instruction verbosity — report.md (2026-05-09) #31273) were internal workflow, test refactor, or hardening changes with no user-facing surface, and were intentionally not documented.
• [model-inventory] Add gemini-flash-lite alias and register 2026-05-10 multiplier gaps #31283 also added new entries to the ET multiplier registry (gemini-3.1-flash-lite, gemini-3.1-pro-preview-customtools, gemini-2.5-computer-use-preview-10-2025); those are surfaced through the auto-generated model-tables.md and required no hand-edits.
• The model-alias-specification.md table has additional pre-existing drift from source (missing gpt-5-pro, deep-research, the gemini/ provider patterns for gemini-flash and gemini-pro, and gpt-5-pro in the large meta-alias). Those gaps predate the last-24h window and are intentionally left for a separate sync PR to keep this change scoped.

Generated by Daily Documentation Updater · ● 7.9M · ◷

• expires on May 11, 2026, 10:59 AM UTC