calcs/concentric-footing/tasks/005_review.md

Task 005 — Reviewer pass on the whole calculation

Goal

Verify engineering correctness, deterministic evidence, and documentation for the concentric footing analysis before marking the project done.

Background

This task is reviewer-local (reasoning) with simplify skill pass. It must not modify source except via review report; if issues are found the orchestrator will send them back to the builder. The reviewer must check ACI 318-19 clause references, units, applicability limits, benchmark tolerances, and that the Typst sheet presents only (no recomputation) per PROJECT_STATE.md.

Files to Modify

No files to modify in this task. The reviewer writes a review note (verbal output) and runs the simplify skill as part of the standard review pass. If STATUS: FAIL, the orchestrator updates TASKS.md and re-delegates.

If the reviewer finds fixable issues, they must be reported as a structured list with file and line references.

Implementation

Reviewer steps (read-only, then run commands via delegation tool if needed — but orchestrator handles command execution; reviewer describes what to run and checks output):

1. Read PROJECT_STATE.md, TASKS.md, calcs/concentric-footing/calc.py, input.yaml, results.json, footing.typ, test_concentric_footing.py, README.md, codemap.md, and CONCENTRIC-FOOTING.pdf.
2. Verify engineering spec. Formula → code-section map (all in calc.py:compute()):
   • Soil bearing q=Ps/Af, qu=Pu/Af: Pressures block → q_psf, qu_psf.
   • One-way shear Vc=2·λ·√fc·B·d, Vu=qu·B·L1: One-way shear block → L1_in, Vu_one_way_kip, Vc_one_way_kip, phiVc_one_way_kip.
   • Two-way shear bo=4(c+d), vc=min(4,2+4/β,2+α·d/bo)·λ·√fc, Vu=qu·(Af-(c+d)²): Two-way shear block → bo_in, vc_psi, Vu_two_way_kip, Vc_two_way_kip, phiVn_two_way_kip.
   • Flexure Mu=qu·B·Lc²/2, a=As·fy/(0.85·fc·B), Mn=As·fy·(d-a/2): Flexure block → Lc_in, Mu_kipft, a_in, beta1, Mn_kipft, phiMn_kipft.
   • Bearing Bn=0.85·fc·A1·min(√(A2/A1),2): Concrete bearing block → A1_in2, A2_in2, sqrt_ratio, Bn_kip, phiBn_kip.
   • Soil bearing q=Ps/Af and qu=Pu/Af correctly use Af=Bf^2 in ft2, Ps/Pu in kip, qa in psf, D/C correct.

• One-way shear Vc=2lambdasqrt(fc)Bd with phi 0.75, L1 = (B-c)/2 - d, Vu = quBL1, correct unit conversion lbf->kip, handles L1<=0 edge.
• Two-way shear bo=4*(c+d), vc=min(4,2+4/beta,2+alphad/bo)lambdasqrt(fc) with alpha 40 interior, beta 1, phi 0.75, Vu=qu(Af - (c+d)^2), bo correction documented.
• Flexure Mu=quBLc^2/2, a=Asfy/(0.85fcB), Mn=Asfy*(d-a/2), phi 0.9, beta1 formula per ACI, rho vs 0.0018.
• Bearing Bn=0.85fcA1*min(sqrt(A2/A1),2), phi 0.65, A1=bp^2, A2=Bf^2, sqrt cap.
• d = Df - cover to centroid, validation, lambda bounds, rebar area from #size.

3. Verify deterministic evidence:
   • Run python -m pytest calcs/concentric-footing/test_concentric_footing.py -v and confirm 12 passed, no skipped.
   • Run python calcs/concentric-footing/calc.py --stdout | python -m json.tool and confirm idempotent vs results.json (diff empty).
   • Run typst compile --root . calcs/concentric-footing/footing.typ calcs/concentric-footing/generated/footing.pdf and confirm PDF exists and size >50KB.
   • Run Typst queries for <concentric-footing-loads> and <concentric-footing-results> and confirm reconciliation within tolerances.
4. Check scope/limitations note in footing.typ mentions square only, interior only, gross pressure, no moment, bo correction, plate clarification, d definition, and that existing calc files were not touched.
5. Run simplify skill pass on the diff (calc.py + footing.typ + test file) for readability, no behavior change.
6. Return STATUS: PASS if all 8 checks below are satisfied, else STATUS: FAIL with enumerated issues.

Acceptance Criteria (all must be true for PASS)

1. All six checks pass for default input; D/C values within expected ranges (soil ~0.82, one-way ~0.05, two-way ~0.11, flexure ~0.12, bearing ~0.22).
2. Corrected punching perimeter bo=92in is used; review notes the PDF's 68in deviation and that the sheet's Scope documents it.
3. Bearing plate vs column clarified (A1=36in2, A2=1296in2, Bn 183.6kip); alternative use of column size would be noted.
4. Pint unit conversions tested (alternate units test passes) and wrong-dimension guards raise ValueError with field name.
5. Effective depth d = Df - cover correctly validated; zero/negative d raises ValueError.
6. Typst sheet presents only (no capacity recomputation), derives loads inline, emits both <concentric-footing-loads> and <concentric-footing-results> metadata, and compiles.
7. README and codemap refreshed and list the new calculation; no existing calc files were modified (git diff or file timestamps check).
8. Pytest 12 passed, PDF compiles, results.json idempotent, simplify pass clean (no behavior change).

Tests

Reviewer (via orchestrator delegation) runs:

python -m pip install -r requirements.txt
python -m pytest calcs/concentric-footing/test_concentric_footing.py -v
python calcs/concentric-footing/calc.py --stdout
typst compile --root . calcs/concentric-footing/footing.typ calcs/concentric-footing/generated/footing.pdf

All succeed; 12 passed.

Dependencies

Task 004 must be DONE.