punkshell/src/tests/modules/AGENTS.md

src/tests/modules — Source Module Test Suites

Purpose

Unit tests for editable source modules under src/modules/, src/modules_tcl8/, and src/modules_tcl9/. These tests use tcltest and run against the uninstalled source tree through src/tests/runtests.tcl.

Ownership

• Agents should add or update tests here when changing module behavior in src/modules/ or Tcl-version-specific module trees.
• Installed-module test packages under src/modules/test/ are packaging artifacts and are not the default place for source-tree behavior tests.
• Do not delete, skip, or weaken existing tests without explicit user direction.

Local Contracts

• The directory hierarchy mirrors module namespace paths, such as src/tests/modules/punk/args/ for punk::args.
• Test files use .test extension and must package require any extra packages explicitly.
• Test files must end with tcltest::cleanupTests so src/tests/runtests.tcl can produce trusted aggregate counts.
• Test files may commonly use line-continuation backslashes, and existing Tcl expand-syntax structures in tests should not be refactored into line continuations.
• Tests must exercise source modules as loaded by src/tests/runtests.tcl, not installed or built output modules.
• testsuites/dev/ may hold known-bug or expected-fail tests; guard them with disabled tcltest constraints so standard source-tree runs skip them by default.

Work Guidance

• Put new module tests under src/tests/modules/<namespacepath>/testsuites/<suite>/ following nearby layout.
• For efficient agent runs, use -report compact -show-passes 0 and include the narrowest namespace path plus file tail when known.
• Prefer try { ... } on error {result options} { ... } for structured error capture in test bodies.
• Focused verification should mirror production pipelines inside tests using Punk pipeline syntax when that parity matters.
• Capture enough stderr or failure context to identify the failing command or assertion.

Verification

• Run targeted module tests with <tcl_interpreter> src/tests/runtests.tcl -report compact -show-passes 0 -include-paths modules/<namespacepath>/***.
• Run a single test by name with <tcl_interpreter> src/tests/runtests.tcl -report compact -show-passes 0 -tcltestoptions {-match <test_name>} -include-paths modules/<namespacepath>/***.
• Run a specific test file by adding its file tail, such as <tcl_interpreter> src/tests/runtests.tcl -report compact -show-passes 0 -include-paths modules/punk/args/*** parsekey.test.
• Use -report markdown+json when both detailed human-readable failure sections and a structured summary are useful.
• If a compact result shows reason=missing-cleanupTests, add the missing tcltest::cleanupTests before trusting observed pass events.

Child DOX Index

• opunk/console/ — ::opunk::Console backend subclass tests (testsuites/console/backends.test, G-001): virtual dispatch of subclass overrides through base-class calls and punk::console::console_spec_resolve (both unchanged), TestConsole determinism + probe-free at_eof, SshConsole capability/eof + the flagship size-via-ANSI-query-over-socket case (a scripted remote terminal answers CSI 6n), TkConsole widget size/eof (gated behind env PUNK_TEST_TK=1 - Tk in the shared testinterp has side effects; also verifiable standalone under a tk-capable kit e.g punk91 src <script>)
• punkboot/utils/ — punkboot::utils tests (testsuites/utils/): the make.tcl helper module. utils.test (punkproject.toml/CHANGELOG version parsing), vcsdirty.test (dirty fossil/git provenance warnings behind the producing-commands gate - git-fixture based), and bootlibrary.test (G-125 boot-precondition predicate vfs_boot_library_report: both tcl-library conventions - tcl_library/ for zipfs-attached kits and lib/tcl<major>.<minor>/ for starkit-style kits - the companion-file requirement that stops the lib/BWidget1.10.1/init.tcl every punkshell kit carries from answering for a tcl library, near-miss reporting, missing/empty trees, and a sweep asserting every assembled src/_bake/*.vfs tree still passes so the gate cannot fail kits that boot today). All three are pure fixture tests - no mint or bake is run; the make.tcl side of the gate is pinned separately in shell/testsuites/punkexe/maketclbootgate.test
• modpod/ — modpod (vendored zip-based .tm wrapper) tests (testsuites/modpod/roundtrip.test, G-111 - the module's first suite): make_zip_modpod wrap emits stub+zip (\x1A separator, PK local header), is_valid_tm_version accept/reject, and child-process load round-trips of per-run generated fixture pods - require from a real-disk module path (stub self-mounts via zipfs, or the vfs::zip fallback on 8.6), the -offsettype file form, the unwrapped #modpod-- redirect form (extracted folder beside the .tm is sourced directly - exact path asserted, no mount signature), a binary payload (dll discovered by a child probe among packages NOT already loaded in a bare child, repo lib_tcl trees offered as auto_path - no committed binaries) loading from the mounted pod in a fresh child, and the tm-residing-on-a-zipfs-path zip-in-zip form (childzipfs-gated: skips on 8.6, the recorded G-034-class limitation). Child spawn probes the kit script subcommand form first, then plain script-file dispatch (native tclsh)
• commandstack/ — commandstack (cooperative command renaming) tests (testsuites/commandstack/commandstack.test, 2026-08-03 - characterisation suite + the G-160 hygiene-pass contract at commandstack 0.6.0 + the 0.7.0 convenience removal forms + the 0.7.1 reload-contract state guards (G-160 follow-ons, 2026-08-04); usage-driven from punk::packagepreference/packagetrace/packagesuppress/punk-auto_execok/punk::nav::fs-cd): record shape as a contract (token first/renamer second dict key order for the lsearch -index 1/-index 3 convention, trailing did_rename 0|1, {implementation {} did_rename 0} no-rename signal), COMMANDSTACKNEXT/COMMANDSTACKNEXT_ORIGINAL delegation + the commandstack::next helper (caller-context parity with the manual uplevel convention pinned), unique+monotonic per-(renamer,command) tokenids (same-renamer re-renames chain and are removable by exact token, third rename succeeds), multi-renamer stacking with removal in any order (bottom-removal re-linking), builtin renames (next_implementor original), remove_rename's three argument forms + unknown-renamer errors, the 0.7.0 convenience removal forms (pop_rename - command form pops topmost-for-renamer and returns the removed record, bare form searches live stacks with a multi-command ambiguity error; remove_renamer - all of a renamer's entries across live stacks with correct re-linking, Rename_stack-parked stacks skipped, removed records returned keyed by command; restore_original - whole-stack unwind to the original returning the record count, deliberately registering stack-evidenced renamers so it survives a known_renamers reset while the renamer-explicit forms keep the gate; all silent with debug off), the token->implementation map get_next_command resolves through (map/stack consistency pinned across rename/remove; parked stacks keep dispatching), channel discipline (silent full cycle with debug off; warnings only under debug), debug argument validation, -renamer misplacement errors, get_stack raw-key-first retrieval of Rename_stack-parked records + Rename_stack 1/0 returns, Delete_stack live-record guard (errors; empty/missing return 1), get_IMPLEMENTOR classification incl builtin-where-cmdtype-exists (dynamic expectation - undetermined on 8.6), lib::split_body round-trip, lib::splitx, show_stack fallback render, the reload contract (a module re-source preserves ALL state - stacks/token counters/token map/known_renamers/debug each info-exists guarded; delegation and exact-token removal keep working after reload), the help overview, and lazy punk::args registration of the PUNKARGS docs. Behavioural tests run in fresh child interps per test (module sourced by path relative to the test file; a ::puts shim captures module output for silence/warning assertions and keeps runner output clean); descriptions are single-line per the tcltestrun banner-parsing style guidance in src/tests/AGENTS.md (a hard contract until G-161 made the parser multi-line tolerant). Green on tclsh90 (9.0.3) and punk86 (8.6)
• punkcheck/ — punkcheck module tests (install, summarize_install_resultdict, installtrack)
• punk/ansi/ — punk::ansi tests (testsuites/ansi/): ansistrip/ansimerge, plus characterization of the ANSI-at-position mechanisms (ansistring.test: INDEX/INDEXCODE/INDEXCHAR/RANGE/INSERT grapheme indexing with SGR-prefix merging, INDEXCOLUMNS/COLUMNINDEX double-wide column mapping, trim/VIEW; extended 2026-08-05 by the G-151 pre-modification coverage survey - VIEW -lf 0/1/2 single-line-ization modes + always-on NUL + silent unknown-option tolerance (load-bearing for overtype's '-nul 1' sites) + C1/zero-width visuals, RANGE end-relative/clamping plus the merged-code-stack re-emission and trailing-code-drop semantics a truncate primitive must respect, COUNT-vs-length combining-accent divergence, NEW object basics, and the KNOWN-DEFICIENCY pin that VIEWCODES/VIEWSTYLE raise invalid-command in a plain punk::ansi interp (bare 'a+'/'a' interactive-alias dependency; pinned in a fresh child interp); plus the TRUNCATE primitive pins landed with G-151 (punk::ansi 0.2.0) - fits-unchanged byte identity incl styled fixtures, grapheme-exact capping with marker budget participation and marker-alone truncation, SGR-reset-before-marker on styled prefixes, CJK/combining-cluster grapheme safety, and strict option errors (deliberately unlike VIEW's pinned silent tolerance)), code splitting invariants (ta.test: detect/detectcode distinction, split_codes/split_codes_single/split_at_codes shapes and round-trip) and single-code/effective-state semantics (codetype.test: is_sgr_reset/has_sgr_leadingreset, has_any/all_effective, sgr_merge, sequence_type classify), grepstr characterization (grepstr.test: return modes incl summarydict (linemap pinned as always-present - the -help says -n-only, reconciliation deferred to the planned hygiene pass), exact highlight SGR wrapping, -n line numbering, invert + empty-highlight strip, -C context/breaks, capture groups, and the tab deficiency: warns once per call on stderr, single-pass tab line survives - the multi-pass mangling is pinned at consumer level in punk/ns corp.test), and untabify characterization (untabify.test: -stops int/list/terminal, -with spaces/unicode/custom-pair, multiline, errors, plus the EXPERIMENTAL -plastic elastic-tabstop mode deliberately pinned-as-interim and retained for possible repl editbuf use). Console queries (get_tabstops/get_size + punk::console::tabwidth) are mocked per the overtype renderline.test pattern - they emit live terminal queries that block/error headless. ANSI codes in these tests are literal escape strings so results are colour-state independent
• punk/args/ — punk::args tests (testsuites/args/): parsing, choices/choicegroups, forms (incl the 2026-08-05 candidacy fencing/fall-through primitive pins ahead of G-168: -regexprefail on optional value slots is form-fatal with no re-landing, unknown-option/dangling-option/option-value-fence failures fall through to sibling forms as successful parses, and the documented per-word limit - a complete option+value+file line double-matches a fenced scriptfile form and a dash-tolerant stdin form), registered-namespace lazy definition loading (docpackages.test, G-169 pre-work: inert registration/scan-vs-load split, id_exists never triggers loading, real_id/usage lazy resolution incl tag-prefixed script-level id families, duplicate-definer last-loaded-wins characterization), the G-168 launch-definition model parity (punkexemodel.test, punkexe moduledoc 0.7.0, id loading updated for the G-169 handover - the script id resolves via app-punkscript-docs from src/lib (auto_path derived from the test file location) while the moduledoc carries the core-owned ids: script/tclsh selection matrices against the real (script)::punkexe ids with the real-side dispatch/app oracle recorded in goals/archive/G-168, the bare-'-e' viable-incomplete verdict, the ./-e fence message, -encoding fall-through statuses, and the sanctioned complete-'-encoding' multipleformmatches LIMIT pins; the formcheck.test punkexe GAPs flipped to discriminated/sanctioned/selection-sound pins in the same arc), the G-151 landing report (parsereport.test, punk::args 0.18.0 parse_report: the canonical flag-like-word-consumed-as-VALUE attribution row, parsed-result vs words+withid entry parity, machine dict shape with declaration-section row order and absent-optional row omission, received xN multiplicity for solos/-multiple opts, type-aware always-marked elision at the default width plus the caller > @cmd -reportvaluewidth > built-in width cascade and never-elided dict returns, VIEW-style single-line-ization of control-bearing values, aliased-optionset storage-key row bridging, bordered-table/tableobject renders, words-form failure errorcode parity with parse, and the registered two-form definition's own render), rendering/indentation characterization, synopsis display characterization (synopsis.test: basic italic argname/<type> styling, longopt --x= alias forms, literal/literalprefix/stringstartswith/stringendswith type-alternates rendering unitalicised, option alternate parenthesization, multi-element clause display incl ?type? members and argname tail-word hints, -typesynopsis value-element lists and option passthrough incl documenter ANSI, and the small-restricted-choice-set literal rule: 1-3 restricted choices render as unitalicised |-joined literals in leader/option/value positions with choicegroups counted, >3 or -choicerestricted 0 falling back to italics, -typesynopsis taking precedence), usage-marking characterization (usagemarking.test: -parsedargs/-badarg/-parsestatus/-scheme marking primitives plus goodchoice highlighting of selected/default-in-effect choice words, asserted by SGR-parameter subset against the live colour arrays; the G-049 nocolour/colour-leak GAP pins flipped 2026-07-10 to scheme-statelessness assertions), the G-049 parse-status structure (parsestatus.test: punk::args::parse_status overall/per-argument statuses, badarg for type/allocation failures, -caller attribution, errorcode -argspecs stripping), -parsekey characterization (parsekey.test: result/received/solos/multis keying, shared-key required satisfaction and defaults, mash-path and prefix-abbreviation keying, plus GAP pins for last-defined-member default precedence, cross-member -multiple value loss, and values/leaders parsekey breakage - desired-behaviour pins disabled behind punkargsKnownBug in testsuites/dev/parsekey-knownbugs.test; the parsekey/optname collision GAP is flipped (G-083, 2026-08-07: -parsekey colliding with a distinct defined arg's name is now a resolve error unless that arg shares the parsekey), and the G-083 argument-relations define-time vocabulary is pinned in relations.test (per-arg -conflicts target validation, @opts -parsekeymode override|error storage and the requires-group/bad-value/requires-parsekey resolve errors), plus its increment-2 parse-time enforcement (optionconflict raise for -conflicts violations and -parsekeymode error co-occurrence, defaults-never-conflict, cross-group conflicts, parse_status invalid classification), and its increment-3 usage/synopsis rendering and lsearch moduledoc adoption (conflicts-with hints appended to per-arg help rows, -parsekeymode error groups annotated 'mutually exclusive' in group headers with override groups unannotated, synopsis one-line form carrying no conflict detail; and the lsearch tclcore moduledoc modelling -sorted/-glob and -bisect/-all incompatibilities via per-arg -conflicts while -glob/-regexp stay last-wins and the 'punk::args fixes required' caveat is dropped))), the G-053 range-valued -multiple occurrence arity (multipleranges.test, punk::args 0.22.0: {min max} range canonicalisation into a per-form MULTIPLE_RANGES companion dict while the stored -multiple boolean keeps its hot-path list-collect meaning so legacy 0/1 stay byte-unchanged, the -optional/range-min reconciliation resolve error, the {0 1}/{2 4}/{1 -1} scalar-vs-list value shapes, parse-time occurrencecount enforcement with min-under-supply suppressed in the G-152 viability probe and classified incomplete (pure exhaustion) while over-max is invalid, the usage-table Multi column showing 0-1/2-4/1+ with the synopsis ?arg? at-most-once ellipsis distinction, and -multipleunique composition), longopt style-distinction and mash-edge characterisation (longopts.test, 2026-08-07 gap-fill: the gnu (--flag= inline-only; spaced/solo usage = badoptionformat) vs plain (--flag spaced-only; inline = = invalidoption) vs mixed (-f|--file|--file= both forms incl longopt prefix abbreviation) definition-grammar distinctions, first-=-only value splitting and empty inline values, the two =-member resolve-time definition errors, single-dash -flag=value never split at the =, and @opts -any 1 adhoc passthrough where --flag=value splits but -flag=value stays a whole adhoc flagname expecting a value; the same gap-fill added to mashopts.test: unknown-letter-in-mash invalidoption rejection and the -any interplay - defined flags still mash while undefined mash-lookalike tokens are adhoc flags, never mash attempts; error-shape pins use the first three -errorcode elements), and tclcore doc/interpreter behavioural parity (tclcoreparity.test, G-054, gated on have_tclcoredocs: 'string is' class choices equal the live-harvested set, per-class docids exist, error-vs-ok agreement across the probe matrix; version-note labels appear when the class is live OR is a declared forward unavailable class - dict labels on every runtime since G-073, unicode keeps the live-only rule; the G-073 forward-class adoption invariants and the USER-SANCTIONED 'di' prefix-strictness exemption (the model is deliberately stricter than real 8.6, full words parity-true); and the G-166 pin that an unavailable class's per-class virtual id LEADS its help with the unavailability statement on runtimes lacking the class while modern runtimes generate it unchanged. All expectations are derived from the running interpreter, green on 8.6/8.7/9.0 through runtests.tcl on each - native Tcl 8.6 has been a supported runner interpreter since 2026-07-21)
• punk/nav/ns/ — punk::nav::ns tests (testsuites/nav/navns.test): the n/ n// n/// navigation state machine (ns/ transitions absolute/relative/glob-no-nav, failed-nav state preservation, quad-colon normalization, v-form content selection, ensemble annotation) and the ::punk::nav::ns::ns_current variable contract the repl/codethread/subshell seeding all consume; display content is covered in punk/ns nslist.test
• punk/repl/ — punk::repl tests (testsuites/repl/): opunk console backend integration (consolebackends.test) and repl current-namespace retention (nscurrent.test: real codethread via repl::init driven by synchronous runscript sends - inscope evaluation of ns_current, retention across submissions, n/-navigation retained, auto-create-with-notice for missing namespaces, the 2026-07-14 stray-namespace seeding fix pinned behaviourally plus a source-text guard on repl::start's inline template; the end-to-end piped subshell session is covered at shell level by shell/testsuites/punkexe/shellnavns.test - which found the first-subshell shared-code-interp asymmetry and the piped-inscope gap recorded there)
• punk/ns/ — punk::ns tests (testsuites/ns/): cmdwhich/cmdinfo/cmd_traverse doc-lookup flow (cmdflow.test, G-040 parity; plus the G-166 availability axis on the flowunavail/flowunavail_nodoc fixtures - the unavailable cmdinfo key present on every result, exact and unique-prefix landings on a -choiceunavailable name resolving that name's virtual docid ATTRIBUTED rather than resolving nothing, the choice traverse attributing without addressing when no virtual docid exists, and cmdtype staying unchanged throughout since availability is a second axis, never a cmdtype value), n/ display machinery characterization (nslist.test: tier A get_ns_dicts classification buckets as the machine contract for display reworks - incl package tail/prefix derivation, alias edge cases, usageinfo scan-dependence; tier B per-element layout-agnostic marking - underline/underdouble/underdotted namespace package styles, command type tag colours, exported/imported markers, the punkargs doc icon; tier C REWORK-flagged pins of the current hardcoded 2-col/4-col layout and nspath subtables, to flip deliberately with the planned punk-tables/width-responsive rework; plus the flipped nslist_types_default pin - bare nslist without -types displays all member types since punk::ns 0.7.1 fixed the braced-literal -types default), corp proc-retrieval and syntax/untabify interplay (corp.test: name edge cases, -ranges/-n line handling, basic-highlight ansistrip equivalence, -untabify spaces/unicode tab-free output, the KNOWN-DEFICIENCY pin for default -untabify none on tabbed bodies - grepstr warns per pass and brace overlays mangle tabbed lines, deterministic under mocked console tabstops - and a ::tcl::CopyDirectory -untabify spaces smoke test; precursor coverage for the planned punk::ns hygiene pass), cmdtrace characterization (cmdtrace.test: -pause 0 non-interactive runs, linedict line-mark keys for flat and 2-word-form nested switches as correct-mark guards, and GAP pins for the upstream nested-switch mismark - core.tcl-lang.org tktview 5d5b1052280c976ea3d4, arm bodies whose split-list index lands on a literal switch-command word report container-relative lines; mark tests gated on have_tclcoredocs because cmdtrace's arm-offset correction parses against the ::switch argdoc; plus the fixed-canary asserting punk::lib::check::has_tclbug_nestedswitch_tracelines still reports the bug - a live behavioural probe, so a fixed Tcl release fails the canary first and triggers the documented flip workflow), cmdhelp usage-rendering integration (cmdhelp.test: scheme selection, goodarg/badarg marking incl type/allocation failures, goodchoice highlighting of supplied/default choice words, alias path, cmdinfo result shape, queried-command failure attribution, and -return dict parse-status returns (G-049 - its GAP pins flipped 2026-07-10); the flipped G-051 pins for pseudo-command cmdtype (doconly) + space-form docid prefixes (real string is pins behind the have_tclcoredocs constraint); the G-166 unavailability marking on the helpstrua fixture and on the real string is dict subject - the -return dict key, the table/string marking with the info scheme suppressed so a valid argument tail no longer renders as a cleanly usable command line, and the -return text leading UNAVAILABLE: line (the two deliberate key-list flips - cmdhelp_cmdinfo_result_shape, cmdhelp_return_dict_valid - landed with it); remaining GAP pins for TclOO undocumented-method fallback (G-052) and synopsis marking absence (G-050); plus the G-150 flag-led form-narrowing + selection-soundness pins against punk::auto_exec::hash - bare-parse selection contract (incl the dash-led-name noformmatch cost), no-word/unknown-flag/ambiguity whole-render fallbacks, and hash runtime-behaviour-unchanged), and name/path primitive characterization (nsprimitives.test: string pins for nsparts/nsprefix/nstail/nsjoin/nsjoinall incl weird colon-run (:::) splitting, the trailing-colon parse ambiguity (::x: + y joins to the same string as ::x + :y and reparses leading-colon-greedy), and prefix/tail/join round-trip and its absolutizing exceptions (the original twin-divergence pins for nsparts1/nsprefix1/nsprefix_orig/nstail1/nstail_orig served as safe-deletion evidence and were removed with the twins in the punk::ns 0.7.0 hygiene pass - divergence record in this file's git history, commit 0c7168a1); plus nseval fq-requirement/create-on-eval/evaluator-proc caching, the native-vs-punk p:::x resolution divergence (native namespace eval reaches child x, nseval creates/reaches literal :x), nseval_ifexists no-create + error propagation on plain and genuinely weird namespaces, nsexists/nschildren/nstree_raw weird-ns and relative-resolution pins, globmatchns */**/? semantics (incl * matching a single inner colon - a formerly stale 'should be fixed' comment above nsglob_as_re was corrected in the 0.7.0 hygiene pass), and nspath_to_absolute/nspath_here_absolute caller-resolution pins; the nsjoinall error-message wart pin ('nsjoin:' prefix) flipped when punk::ns 0.7.1 fixed it)
• punk/mix/ — punk::mix::cli tests (prune helpers, punkcheck virtual sources), punk::mix::commandset::repo fossil move/rename characterization tests (testsuites/repo/, FOSSIL_HOME-isolated; GAP-marked tests pin behaviour G-022 will change), punk::mix::commandset::loadedlib tests (testsuites/loadedlib/libsearch.test: 'dev lib.search' match semantics via -return list — wrap-glob default, =exact prefix, case rules, explicit globs, version aggregation — plus the loadedlib 0.2.0 contract: deep discovery by default (deep .tm modules found without -refresh, registration persists), -refresh = genuine re-scan (epoch incr + rediscovery picks up .tm files added to already-scanned dirs), and highlight working without the shell-global a+ alias; shared provisioned child interp sourcing the source-tree libunknown directly — see the file's ORDERING NOTE), and the MULTISHELL polyglot build machinery (testsuites/scriptwrap/, split 2026-07-19 per G-092 so no single file dominates the -jobs parallel floor - tests moved verbatim: multishell.test = scriptset wrap via the punk.multishell.cmd template with structure/LF-only pins plus platform-gated execution smoke (cmd.exe→powershell payload on windows, sh payload on unix or via the wsllinux capability constraint from src/tests/testsupport/wslprobe.tcl - staged to the WSL distro's native filesystem, G-059); multishell_wrapverify.test = checkfile 512-byte label validation of a fresh wrap; multishell_wrapdeterminism.test = byte-identical re-wrap pin; runtimecmd_checkfile.test = checkfile + LF contract of the committed bin/runtime.cmd; runtimecmd_roundtrip.test = the runtime scriptset round-trip byte-identity pin)
• punk/lib/ — punk::lib tests (testsuites/lib/): range/index/parse/compat/interp_sync utilities, G-058 static-baseline seeding (staticseed.test: interp_sync_package_paths/snapshot_package_paths propagate a simulated ::punkboot static baseline and seed load {} <prefix> ifneeded mappings; no-op without a baseline), and the repl command-completeness engine (commandcomplete.test: punk::lib::system::incomplete pending-opener stacks - the info-complete quoting quirk progression (set x "{*}{" standalone vs in-proc-body), single openers, tabs, escapes, incomplete<->info-complete parity property; pre-repl-refactor characterization, see goals/G-044 detail preserve-list)
• punk/packagepreference/ — punk::packagepreference tests (testsuites/packagepreference/): G-058 static-vs-bundled policy (staticpolicy.test: require of a baseline package triggers the index scan before resolution so a newer bundled copy wins, static beats older bundled, exact requires of bundled versions stay reachable, missing static mappings get seeded), and the install re-install guard (installguard.test, G-160: second install returns 0 silently via the record renamer key - the historical rename read threw; uninstall/re-install cycle gets the next unique tokenid; fresh child interps source the SOURCE-TREE commandstack + packagepreference by path so no bootsupport/kit snapshot can answer)
• punk/zip/ — punk::zip tests (testsuites/zip/zipreader.test, G-124 - the module's first tests): the dependency-free reader over the three archive shapes (bare zip, executable-prefixed with archive-relative offsets, executable-prefixed with file-relative offsets) - archive_info's offsetstyle/dataoffset derivation incl. the nozip verdict for a plain binary carrying a stray PK\5\6, members' per-entry introspection (classification, sizes, method, crc, mtime, stored attributes) and glob/exclude selection, unzip's byte-identical CRC-verified extraction incl. the >2MB streaming path and partial extraction; the named refusals (encrypted, unknown compression method, zip64, crc mismatch, path escape) each asserting that no partial output is left behind; and the mkzip->read round trip - the first coverage that what punkshell WRITES is readable. Runs entirely on stock Tcl (verified on msys2's tclsh8.6 with no zipfs, no vfs::zip and no tcllib); two zipfs-gated tests cross-check against a real zipfs mount and record the answer to the standing directory-classification question (zipfs keys on the trailing slash, not the stored permission bits - so punk::zip::mkzip directories mount as directories). The real-runtime pin self-gates on bin/runtime/win32-x86_64/tclsh90b4_piperepl.exe, which is untracked
• punk/libunknown/ — .tm same-version shadowing pin-tests (testsuites/shadowing/): tcl:™️:add prepend rule, head-of-tm-list wins exact-version ties, version beats order, punk::libunknown parity — shipped behaviour depends on these (runtests tm ordering, punk_main package-mode precedence, G-033); mixed .tm/pkgIndex.tcl characterization is goal G-035; discovery/epoch-cache characterization (testsuites/discovery/discovery.test): sibling .tm registration happens only at the requested namespace depth (deeper modules invisible to 'package names' until requested), register_all_tm deep discovery (all-depths registration, per-epoch cached no-op, head-of-tm-list precedence parity), 'package epoch' command shape, trace-driven epoch increments on tm/auto_path changes, the epoch-index short-circuit (a .tm added to an already-scanned dir needs 'package epoch incr'), and the pkgIndex.tcl sourcing-scope contract (source_pkgindex, fixed 2026-07-11 - formerly a GAP pin of the global-'dir' clobber): user global 'dir' survives pkg-unknown fallthrough, index scripts see $dir and reach the real ::auto_path (tcllib extension pattern), and their stray unqualified sets no longer leak to ::. Suite conventions learned the hard way: child probes source the SOURCE-TREE libunknown directly by path — 'package require punk::libunknown' tie-breaks the same-version copies (src/modules vs bootsupport) by tm path order, which favours bootsupport in the testinterp and machine env paths in children; and probe scripts must avoid global variable names the handlers use (notably 'dir') — that clobbering misdirected a fixture write into the real Tcl install during test development