[squid-protocol]

Part of #5154. Cross-repo: none (engine only).

What changed

Box on #5154	Result
DFHJCT JFILEID= / JTYPE=	Open, kept ASSUMED by name. IBM TS 5.x has no JCT page; the only IBM statement found is that JFILEID is "supported for compatibility purposes only" (TS 5.3 recov_journaling). Header + X62 say so.
DFHSIT SUFFIX=	Ticked: confirmed (TS 5.4 dfha2_suffix: 1-2 chars, not NO/DY, SIT only). Row is now named DFHSIT+suffix, macro_table the same; an invalid suffix names nothing.
SIT parameters as rows	Ticked: refused with reason (start-up SIT=/SYSIN overrides); kept in attributes.
DFHTCT LINE / SDSCI	Ticked: refused with the IBM reason (TS 5.5 tct/seq.html: SDSCI is a DD, LINE joins two).
DFHTCT GPENTRY	Open: no IBM page retrievable; stays refused.
DFHMCT TYPE=RECORD	Ticked: refused (TS 5.5 mct/record.html: selects performance fields, names no resource).
JCL SYSIN as separate member	Ticked: GalaxyIR.job_macro_tables() joins the member's macro rows to the step (member missing/ambiguous reported, no rows). Override DDs (//STEP.SYSIN) not read.
PCT / PPT / SDSCI ASSUMED operands	Open: no IBM page for DFHPCT, DFHPPT, PARTITIONSET/PARTSET or RECFORM=UNDEFINED; stay ASSUMED. The CICS/VS and ESA 3.x macro manuals are still the references to confirm against.

Tests / gates

New unit tests (SIT suffix class, SIT params never rows, SDSCI/LINE, MCT RECORD) and a real-scan join test with a new jcl_member fixture (found, missing, case-blind, no member, non-assembler step). pr_gates --fast 6/6, --ratchets 8/8, crucible_check.py both legs PASS (no golden drift).

🤖 Generated with Claude Code

[squid-protocol]

No description provided.

[github-actions]

🐦‍⬛ Muninn Security Scan

✅ No security issues found.

🐦‍⬛ Powered by Muninn · Skald Lab

[squid-protocol]

No description provided.