The tcl-whatis and tcl-nslist skill descriptions contained ': ' mid-string,
which is illegal in a plain YAML scalar and broke frontmatter parsing in
strict harnesses. Wrap both description values in double quotes (all four
copies: .agents/skills and .claude/skills).
Assisted-by: harness=claude; primary-model=claude-fable-5; api-location=anthropic.com
The skills existed but sessions never leaned on them - grep+read wins
the tool-selection contest when descriptions are phrased as information
needs ("use when you need to know how to call...") rather than task
triggers, and grep 'succeeds' without revealing which of the repo's
coexisting module copies is loaded. Changes to convert them from
available options into workflow steps:
- SKILL.md descriptions rewritten task-first ("Use BEFORE reviewing,
testing, editing..." / "Use BEFORE writing any new Tcl helper..."),
with run-it-first wording, the multiple-module-copies rationale, and
cost stated up front (one sub-1s side-effect-free invocation).
- Invocations shortened to lib: script-path form (verified working,
~0.45s each): punk91 src script lib:developer/whatis|nslist.
- New caveat in both: prefer the Bash tool over PowerShell (attached
console can flip colour/terminal detection in punkshell children).
- AGENTS.md Repo-wide Notes entry converted from reference note to
DIRECTIVE: whatis/nslist first, grep reserved for full-text searches.
- .claude/settings.json: allow entries for both scripts (Bash +
PowerShell, punk91/bin/punk91/./bin/punk91 prefixes) so first use is
prompt-free.
.claude/skills copies re-synced from canonical .agents/skills
(diff-verified byte-identical).
Assisted-by: harness=claude; primary-model=claude-fable-5; api-location=anthropic.com
New scriptlib/developer/nslist.tcl backing the new cross-framework tcl-nslist
agent skill (.agents/skills/ canonical, .claude/skills/ synced copy): the
discovery counterpart of whatis.tcl - lists what a namespace/package actually
exposes at runtime via punk::ns::get_ns_dicts (the n// machinery): child
namespaces with loadable-package markers, commands classified by kind,
punk::args-documented commands, -synopsis whole-namespace usage lines,
-pathcommands namespace-path resolution. Auto-package-requires unloaded
packages and follows ensemble commands to their implementation namespace.
The tcl-whatis skill/script cross-reference it (whatis redirects with a hint
when the queried name is a namespace, not a command).
punk::ns 0.9.1: get_ns_dicts packagetails/packageprefixes initialization moved
inside the per-namespace loop - with multiple matched namespaces (path globs,
trailing **) every dict after the first carried the union of all earlier
namespaces' entries, mis-marking non-package children in nslist/n// underline
styling and in the returned data; a child with several deeper packages below
it was appended to packageprefixes once per matching package (now once).
Pinned in nslist.test (nslist_classification_multins_package_buckets_reset).
Documentation-only PUNKARGS added for get_ns_dicts (returned dict keys
documented) and nslist_dict; parsing stays manual.
Root AGENTS.md: introspection note + scriptlib/developer inventory updated.
Project version 0.15.1 + CHANGELOG entry. punk::ns suite 102/102 pass.
Assisted-by: harness=claude; primary-model=claude-fable-5; api-location=anthropic.com
punk::ns 0.9.0: cmdhelp (i) gains -return text - plain-text argument
documentation for machine/LLM consumption: no ANSI, no table layout, the
argument section of every command form rendered in one output under a common
header ('i -return text after' shows all six forms), G-041 form selection
still honoured when argument words are supplied, no advisory-parse error
preamble (-return dict remains the parse-status surface). Implementation
delegates resolution to the -return dict recursion and renders per-form via
the string renderer.
punk::args 0.12.6 (user direction; string-form layout explicitly not frozen -
no external consumers yet):
- arg_error renders the error message + per-form match detail exactly once,
BELOW the usage display (was above and below); bottom append suppressed for
empty messages.
- string renderer argument rows are single-line
'name TYPE:type(constraints) ?DEFAULT:'val'? ?MULTI:yes?' - type
constraints parenthesize onto the base type instead of line-breaking
mid-row; empty DEFAULT/MULTI labels and whitespace-only help blocks omitted.
- choices render as markdown-subset bullets '- `choice` - label' with label
continuations at a fixed deeper indent - choice names are now
machine-distinguishable from label text without ANSI. Structural contract:
column-0 lines bearing TYPE: are argument rows, indented lines are
documentation, backticked bullets are choices. Table cells unchanged.
New scriptlib/developer/whatis.tcl (runtime command introspection for
agents): resolved name, subcommand-chain resolution via punk::ns::cmdinfo
('whatis string is'), kind, providing package + loaded version, origin file,
proc definition file:line, punk::args synopsis; -body appends the runtime
body with file-relative line numbers, -doc the full plain-text argdoc via
cmdhelp -return text (with fallback for older punk::ns). Backing the
cross-framework tcl-whatis agent skill: canonical .agents/skills/tcl-whatis/
(Agent Skills standard - pi/OpenCode/Codex) with byte-identical copy in
.claude/skills/ for Claude Code; .gitignore gains !.claude/skills/ with the
fossil ignore-glob comment + exception-set docs updated per the dual-VCS
contract (fossil adds staged). Root AGENTS.md: .agents/skills index entry,
whatis pointer in Repo-wide Notes, scriptlib/developer example.
Tests: new pins cmdhelp_return_text_all_forms/_argword_form_selection/
_single_form_plain_marker (cmdhelp.test), rendering_stringmode_singleline_argrows
+ rendering_stringmode_choice_bullets (rendering.test); deliberate-change pin
updates: errorselection.test message matches gained leading * and cmdhelp.test
message pins sample the last line (message placement). Full source-tree suite
963/979 pass, sole failure the pre-existing exec-14.3 baseline; make.tcl
modules clean.
Project version 0.12.42 -> 0.13.0 with CHANGELOG.md entry.
Assisted-by: harness=claude; primary-model=claude-fable-5; api-location=anthropic.com
Stock tclsh treats any leading-dash arg other than -encoding as argv for an
interactive/stdin session: on a console it hangs at a prompt, with piped stdin
it exits 0 silently ignoring the supposed one-liner. Agents habitually reach
for -e (perl/python/node reflex), wasting time waiting on hung processes.
- root AGENTS.md User Preferences: document the misparse, the correct patterns
(temp .tcl file or script via stdin) and the defensive stdin-redirect habit
- .claude/settings.json: PreToolUse hook denying Bash/PowerShell invocations
of tclsh with -e, anchored to command position (start/pipe/semicolon/paren)
so prose mentions in commit messages or grep args do not false-positive;
corrective guidance returned to the agent
- .gitignore: narrow the .claude ignore so shared settings.json is tracked
while session-local files stay ignored
- .fossil-settings: ignore-glob hand-derived per the no-negation rule
(.claude/* wholesale + explicit fossil add of settings.json), exception set
in AGENTS.md updated; fossil extras under-ignore check verified empty
Assisted-by: harness=claude; primary-model=claude-fable-5; api-location=anthropic.com
Julian Noble