review-perf

Name: review-perf
Author: gtrabanco

32installs2forks

How to install review-perf?

gtrabanco/agentic-workflow/review-perf

$npx -y skills add gtrabanco/agentic-workflow --skill review-perf

Installs into the current project.

Summary

Internal performance review pass of the agentic-workflow review pack — composed in-turn by review-change and product-audit; not a menu entry. Checks the changed paths for algorithmic and resource regressions: N+1s, hot-path allocations, asset weight, and leaks. Findings only; nev

Files · 1

View on GitHub

SKILL.md

1	# Review Performance (internal)
2
3	Composed by `review-change` / `product-audit` within their conversation — on any
4	agent, follow this file inline as the routed step. **Findings only; never edits,
5	never refactors.**
6
7	## Scope
8
9	The diff or path/glob the caller passes; default the current change vs the
10	default branch. State the scope at the top of the returned table.
11
12	## Checklist (evaluate EVERY item — none is optional; n/a must be stated)
13
14	- ✓ No N+1 query/IO pattern on changed data-access paths (loop bodies issuing
15	queries/requests — cite)
16	- ✓ Algorithmic complexity of new code paths stated where input can grow (an
17	unbounded O(n²) on user data is a major finding)
18	- ✓ No blocking IO/sync work added on hot/request paths where the platform is
19	async
20	- ✓ Resources opened by the change are closed/released on ALL paths including
21	errors (files, connections, subscriptions, timers)
22	- ✓ No unbounded growth introduced (caches without eviction, arrays that only
23	append, listeners never removed)
24	- ✓ Web only: bundle/asset impact of the change stated (new dep size, image
25	weight, lazy-loading where the project uses it) — n/a otherwise
26	- ✓ Repeated computation hoisted where a loop recomputes an invariant (cite)
27	- ✓ Pagination/limits on any new listing that reads user-scaled data
28	- ✓ No premature optimization either: complexity added for speed without a
29	cited measurement is a finding (the repo forbids overengineering)
30	- ✓ Measured evidence when declared — the project's agent guide declares a
31	`Performance commands` block with a `bench` command AND the diff touches
32	paths its benchmarks cover: RUN the benchmark on the base branch and on the
33	change, and cite both numbers in Evidence as
34	`<cmd> → base <x> / change <y> (<±z%>)`. A regression beyond the noise band
35	(the project's declared band, else ±5%) is a major finding; a delta
36	inside the band is no finding. The declared command failing (non-zero exit)
37	is itself a finding (the gate can't measure) — never silently skipped.
38	- ✓ No declared perf commands → state exactly
39	`n/a — no declared perf commands` for the item above (never skip it
40	silently), and if the diff adds algorithmic code on input that can grow,
41	add a minor finding recommending the project adopt the tooling via
42	`init-workspace`'s Performance tooling round.
43
44	## Return exactly
45
46	```
47	REVIEW PERF — scope: <scope>
48
49	\| # \| Finding \| Sev \| Evidence \| Suggested fix \|
50	\|---\|---------\|-----\|----------\|---------------\|
51	\| 1 \| <what> \| critical\|major\|minor \| <file:line — or, for measured findings, `<cmd> → base <x> / change <y> (<±z%>)`> \| <smallest action> \|
52
53	Checklist: <n> evaluated, <n> pass, <n> findings, <n> n/a (<which + why>)
54	Summary: <1-2 sentences>
55	Decision: PASS \| FAIL
56	```
57
58	FAIL if any critical or major finding is open; PASS otherwise. Minor findings
59	never block — they route to the caller's triage step.
60
61	## Done when
62
63	- Every checklist item was evaluated with evidence (file:line or command output)
64	or explicitly marked n/a with the reason.
65	- The fixed-format block above is returned — nothing more, nothing less — and
66	no code was changed.

People who liked this also viewed

6 picks

Type

review-perf

32installs2forks

View on GitHub

How to install review-perf?

gtrabanco/agentic-workflow/review-perf

$npx -y skills add gtrabanco/agentic-workflow --skill review-perf

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Review Performance (internal)
2
3	Composed by `review-change` / `product-audit` within their conversation — on any
4	agent, follow this file inline as the routed step. **Findings only; never edits,
5	never refactors.**
6
7	## Scope
8
9	The diff or path/glob the caller passes; default the current change vs the
10	default branch. State the scope at the top of the returned table.
11
12	## Checklist (evaluate EVERY item — none is optional; n/a must be stated)
13
14	- ✓ No N+1 query/IO pattern on changed data-access paths (loop bodies issuing
15	queries/requests — cite)
16	- ✓ Algorithmic complexity of new code paths stated where input can grow (an
17	unbounded O(n²) on user data is a major finding)
18	- ✓ No blocking IO/sync work added on hot/request paths where the platform is
19	async
20	- ✓ Resources opened by the change are closed/released on ALL paths including
21	errors (files, connections, subscriptions, timers)
22	- ✓ No unbounded growth introduced (caches without eviction, arrays that only
23	append, listeners never removed)
24	- ✓ Web only: bundle/asset impact of the change stated (new dep size, image
25	weight, lazy-loading where the project uses it) — n/a otherwise
26	- ✓ Repeated computation hoisted where a loop recomputes an invariant (cite)
27	- ✓ Pagination/limits on any new listing that reads user-scaled data
28	- ✓ No premature optimization either: complexity added for speed without a
29	cited measurement is a finding (the repo forbids overengineering)
30	- ✓ Measured evidence when declared — the project's agent guide declares a
31	`Performance commands` block with a `bench` command AND the diff touches
32	paths its benchmarks cover: RUN the benchmark on the base branch and on the
33	change, and cite both numbers in Evidence as
34	`<cmd> → base <x> / change <y> (<±z%>)`. A regression beyond the noise band
35	(the project's declared band, else ±5%) is a major finding; a delta
36	inside the band is no finding. The declared command failing (non-zero exit)
37	is itself a finding (the gate can't measure) — never silently skipped.
38	- ✓ No declared perf commands → state exactly
39	`n/a — no declared perf commands` for the item above (never skip it
40	silently), and if the diff adds algorithmic code on input that can grow,
41	add a minor finding recommending the project adopt the tooling via
42	`init-workspace`'s Performance tooling round.
43
44	## Return exactly
45
46	```
47	REVIEW PERF — scope: <scope>
48
49	\| # \| Finding \| Sev \| Evidence \| Suggested fix \|
50	\|---\|---------\|-----\|----------\|---------------\|
51	\| 1 \| <what> \| critical\|major\|minor \| <file:line — or, for measured findings, `<cmd> → base <x> / change <y> (<±z%>)`> \| <smallest action> \|
52
53	Checklist: <n> evaluated, <n> pass, <n> findings, <n> n/a (<which + why>)
54	Summary: <1-2 sentences>
55	Decision: PASS \| FAIL
56	```
57
58	FAIL if any critical or major finding is open; PASS otherwise. Minor findings
59	never block — they route to the caller's triage step.
60
61	## Done when
62
63	- Every checklist item was evaluated with evidence (file:line or command output)
64	or explicitly marked n/a with the reason.
65	- The fixed-format block above is returned — nothing more, nothing less — and
66	no code was changed.

People who liked this also viewed

6 picks

Type