docs: drop the local PR template, inheriting the org default

[lesnik512]

Why

This repo's .github/PULL_REQUEST_TEMPLATE.md does not reach the party writing the
PR. gh pr create reads the template only in its interactive editor path or when
passed -T/--template; --body and --body-file skip it, and gh 2.98.0's help
says so outright. Every PR here is opened by an agent through gh, so the file has
never been applied.

The measurement is in modern-python/.github#53. The short version: the convention
checklist appears in 0 of the last 20 merged PRs in faststream-outbox, 0 of 11 in
chat-app, and none in modern-di since #432. Meanwhile the
Why/Design/Non-goals/Verification shape spread widely, including into
modern-python/.github, whose own template said Summary/Changes/Checklist. The
shape tracks AGENTS.md, which is loaded every turn. It does not track the template.

The checklist also duplicated AGENTS.md: its four routes restate the "Where a fact
goes" table and the two-things-outlive-the-PR paragraph. By the admission check, a
fact derivable from a file the agent already loads does not get written.

Design

Delete the template. The repo falls back to the org default, which is a plain
lint/type/test/build checklist true under either convention (modern-python/.github#53
removes its one pre-convention line).

AGENTS.md named the template as the carrier of the body shape. That paragraph is
dropped rather than rewritten: the body shape is expected to come from the agent-skills
convention, so AGENTS.md stops asserting it. What remains in ## Workflow is the
routing that has no other home — what outlives the PR, and where it goes.

Non-goals

• The per-section guidance is dropped, not relocated, and so is the shape
  statement itself. Nothing about PR-body structure moves into AGENTS.md.
• The ADR revisit-trigger clause is not added. It was in some templates and not
  others, and is not being propagated.
• No drift check. With no local template, there is nothing left to silently
  override the default, which was the only reason to keep a local copy.

Verification

Markdown only; no code, config, or CI touched. The deleted file was referenced from
exactly one place, the AGENTS.md sentence updated in the same commit.

Part of modern-python/.github#53.

[github-actions]

Benchmark

Details

Benchmark suite	Current: 5554c30	Previous: 7bea20c	Ratio
benchmarks/test_guard_by_type.py::test_g16_resolve_by_type	3072502.8325441857 iter/sec (stddev: 9.359831152244764e-9)	3078506.4712193278 iter/sec (stddev: 8.766975887809512e-9)	1.00
benchmarks/test_guard_by_type.py::test_g17_resolve_by_type_large_registry	3059674.6085059824 iter/sec (stddev: 8.772582306308739e-9)	2902284.9137836876 iter/sec (stddev: 1.2969850165637569e-8)	0.95
benchmarks/test_guard_cold.py::test_g8_cold_first_resolve	26608.128952099916 iter/sec (stddev: 0.000003926751143294333)	26526.18795384373 iter/sec (stddev: 0.000003842309461840326)	1.00
benchmarks/test_guard_cold.py::test_g8b_cold_first_resolve_cached	20673.92684020106 iter/sec (stddev: 0.000003008344135418888)	20588.53258222596 iter/sec (stddev: 0.0000028543896748858607)	1.00
benchmarks/test_guard_concurrency.py::test_g14_concurrent_cached_hit[1]	430.8484750721754 iter/sec (stddev: 0.00005217489079380822)	415.167879170779 iter/sec (stddev: 0.00006202680447662843)	0.96
benchmarks/test_guard_concurrency.py::test_g14_concurrent_cached_hit[2]	380.1300565250233 iter/sec (stddev: 0.00029018890602242596)	402.2270751308951 iter/sec (stddev: 0.000049783484680285226)	1.06
benchmarks/test_guard_concurrency.py::test_g14_concurrent_cached_hit[4]	346.94536289431704 iter/sec (stddev: 0.0001587389393746127)	369.39793207318775 iter/sec (stddev: 0.00003147853469498484)	1.06
benchmarks/test_guard_concurrency.py::test_g15_concurrent_first_resolve[1]	2444.264804561249 iter/sec (stddev: 0.000027934469058579786)	2426.430034713546 iter/sec (stddev: 0.000025542094061477845)	0.99
benchmarks/test_guard_concurrency.py::test_g15_concurrent_first_resolve[2]	1847.8082315080703 iter/sec (stddev: 0.000028752941817531206)	1848.1630268279373 iter/sec (stddev: 0.000033221485202946756)	1.00
benchmarks/test_guard_concurrency.py::test_g15_concurrent_first_resolve[4]	1247.1563017540743 iter/sec (stddev: 0.000029981334621972336)	1235.6742619250645 iter/sec (stddev: 0.000035573218994515115)	0.99
benchmarks/test_guard_lifecycle.py::test_g6_build_child_container	746397.2431368113 iter/sec (stddev: 3.362522249737058e-8)	739316.0649963923 iter/sec (stddev: 4.315238358194036e-8)	0.99
benchmarks/test_guard_lifecycle.py::test_g6b_build_child_container_auto_scope	693197.5553423208 iter/sec (stddev: 3.277984744956204e-8)	657351.4812517941 iter/sec (stddev: 5.989090692079666e-8)	0.95
benchmarks/test_guard_lifecycle.py::test_g7_request_lifecycle_batch	2228.259199372679 iter/sec (stddev: 0.000009640099470545749)	2216.2969793986845 iter/sec (stddev: 0.000013652670555895106)	0.99
benchmarks/test_guard_lifecycle.py::test_g7c_event_loop_floor_control	61230.20304470377 iter/sec (stddev: 0.0000013810781561153216)	61348.297172610066 iter/sec (stddev: 0.0000014272120369219915)	1.00
benchmarks/test_guard_lifecycle.py::test_g13_teardown_at_scale	44205.71755753592 iter/sec (stddev: 0.0000016621981609970688)	44039.90084203932 iter/sec (stddev: 0.0000016961515478016748)	1.00
benchmarks/test_guard_resolve.py::test_g1_transient_resolve	2099365.2674071784 iter/sec (stddev: 2.10847016010641e-8)	2072055.4162932062 iter/sec (stddev: 3.279811492684218e-8)	0.99
benchmarks/test_guard_resolve.py::test_g2_cached_resolve	3307965.5380044817 iter/sec (stddev: 8.790548950740878e-9)	3353382.8741951007 iter/sec (stddev: 7.459746786386793e-9)	1.01
benchmarks/test_guard_resolve.py::test_g3_deep_chain	755948.0925233179 iter/sec (stddev: 5.5101619712632417e-8)	766752.271989852 iter/sec (stddev: 3.4366223902824446e-8)	1.01
benchmarks/test_guard_resolve.py::test_g4_wide_resolve	412566.06085217703 iter/sec (stddev: 9.573597705937111e-8)	411006.9046441213 iter/sec (stddev: 1.0341918503424523e-7)	1.00
benchmarks/test_guard_resolve.py::test_g5_cross_scope	1662912.5722450528 iter/sec (stddev: 8.264550562803531e-8)	1670319.0977857038 iter/sec (stddev: 2.405138182181244e-8)	1.00
benchmarks/test_guard_resolve.py::test_g9_context_resolve	834795.477611472 iter/sec (stddev: 1.3845835484141966e-7)	879367.3918940694 iter/sec (stddev: 5.977952127212956e-8)	1.05
benchmarks/test_guard_resolve.py::test_g12_override_active_resolve	425876.9023681843 iter/sec (stddev: 4.6449147503354515e-7)	501610.772508836 iter/sec (stddev: 6.58084353100994e-8)	1.18
benchmarks/test_guard_resolve.py::test_g18_alias_hop	2247176.6697042338 iter/sec (stddev: 1.8514061312526047e-8)	2288530.721617111 iter/sec (stddev: 9.141403052020257e-9)	1.02
benchmarks/test_guard_validate.py::test_g10_validate_deep_chain	29958.082650792214 iter/sec (stddev: 0.0000027404979536528853)	30550.40215508519 iter/sec (stddev: 0.000003209223020948812)	1.02
benchmarks/test_guard_validate.py::test_g11_validate_wide	18237.620769677564 iter/sec (stddev: 0.0000028212191065610325)	18181.870083005833 iter/sec (stddev: 0.000003477050944462194)	1.00

This comment was automatically generated by workflow using github-action-benchmark.