Skip to content

docs(aws-serverless): update LMI skill for region expansion and scheduled scaling#1

Open
sbbhimji wants to merge 3 commits into
mainfrom
lmi-scheduled-scaling-region-expansion
Open

docs(aws-serverless): update LMI skill for region expansion and scheduled scaling#1
sbbhimji wants to merge 3 commits into
mainfrom
lmi-scheduled-scaling-region-expansion

Conversation

@sbbhimji

Copy link
Copy Markdown
Owner

Summary

Updates the aws-lambda-managed-instances skill for two recent Lambda Managed Instances launches:

  1. Region expansion (announcement) — LMI is now available in all commercial AWS Regions except Israel (Tel Aviv), Middle East (Bahrain), Middle East (UAE), and Asia Pacific (Auckland). Replaces the previous fixed five-region list and updates the "Unsupported Region" handling to name the excluded regions and suggest the nearest supported one.
  2. Scheduled scaling (announcement) — documents using Amazon EventBridge Scheduler to adjust MinExecutionEnvironments/MaxExecutionEnvironments on a one-time or recurring schedule for predictable traffic (scale up before peak, scale down or to zero when idle).

Files changed

  • SKILL.md: regional availability + unsupported-region scenario; scheduled scaling trigger and scheduled-scaling tag in frontmatter; scheduled-scaling note in Step 3; two Operations best practices; reference-file pointers; Scaling docs link in Resources.
  • references/configuration-guide.md: new "Scheduled Scaling (Predictable Traffic)" section — behavior notes (still scales on CPU/concurrency between bounds; Min=Max=0 deactivates and needs explicit reactivation) and a common-patterns table.
  • references/infrastructure-setup.md: new "Scheduled Scaling (EventBridge Scheduler)" section — scheduler IAM trust/permissions policies, create-schedule examples targeting the PutFunctionScalingConfig universal target, and a manual put-function-scaling-config override.

Testing

  • dprint check — clean
  • markdownlint-cli2 (incl. custom skill-length + frontmatter rules) — 0 errors
  • lint:cross-refs — 0 errors
  • validate:refs — 0 broken links
  • validate:size — SKILL.md 225 lines / 1566 words, description 626 chars (all under limits)

Documentation-only change; content verified against the LMI scaling documentation. Security scanners (gitleaks/checkov/etc.) were not run in this environment — the change adds only docs with placeholder identifiers (<account-id>), no real secrets or IaC files.

sbbhimji and others added 3 commits June 19, 2026 15:53
…uled scaling

- Region availability: now all commercial AWS Regions except Israel (Tel Aviv), Middle East (Bahrain), Middle East (UAE), and Asia Pacific (Auckland); update Unsupported Region handling.
- Scheduled scaling: document EventBridge Scheduler-based scheduled scaling of Min/Max execution environments for predictable traffic (SKILL.md, configuration-guide.md, infrastructure-setup.md).
…e JSONB" (awslabs#200)

* fix(dsql): reframe JSON/JSONB/array storage as a choice, not "MUST use JSONB"

The native JSON/JSONB rollout left this skill prescribing a single answer.
Two drift directions existed: some files said "MUST serialize arrays as
JSONB" / "prefer JSONB over JSON" (over-correction), others still said
"store as TEXT" / "SET -> TEXT" (stale pre-JSON framing).

Bring all nine steering files in line with the already-reviewed wording in
awslabs/mcp: serializing is the MUST (DSQL has no array column type); the
format is a choice — PREFER JSONB for queryable values, MAY use TEXT when
opaque, JSON when writes dominate or byte-exact preservation matters; keep
existing JSON columns as JSON when migrating. ASK the user.

Files: SKILL.md, references/development-guide.md, references/onboarding.md,
references/troubleshooting.md, references/examples/{patterns,schema}.md,
references/mysql-migrations/{type-mapping,full-example,ddl-type-alternatives}.md.
Bump databases-on-aws to 1.3.3.

* ci: exclude broken registry rule bbp-pattern-inject from semgrep

The community registry rule bbp-pattern-inject fails to parse
("Invalid pattern for Python: Stdlib.Parsing.Parse_error"), making
semgrep exit 2 and fail every PR regardless of findings. Exclude it
the same way other unwanted registry rules already are.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants