wzyn20051216/matlab-agent-skills / skills/matlab-runner
matlab-runner
VerifiedRun MATLAB tasks from Codex with reliable MCP-based execution, logs, artifacts, toolbox inventory, smoke tests, and reproducible validation. Use whenever MATLAB code, .m/.mlx scripts, MATLAB tests, local toolbox probing, or command-line automation is needed.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 13.5 KB |
| references/acceptance-levels.md | markdown | 3.0 KB |
| references/failure-classes.md | markdown | 2.5 KB |
| references/mcp-readiness-ladder.md | markdown | 6.4 KB |
| references/toolbox-feature-map.md | markdown | 1.1 KB |
| scripts/mas_dep_probe.m | other | 11.4 KB |
| scripts/mas_precheck.m | other | 7.5 KB |
| scripts/mas_product_map.m | other | 6.2 KB |
| scripts/mas_toolbox_map.m | other | 2.8 KB |
| tests/test_dep_probe.m | other | 6.2 KB |
| tests/test_precheck.m | other | 3.6 KB |
tests/test_precheck.m
View on GitHubfunction res = test_precheck(verbose)
%TEST_PRECHECK Pins each pre-check against a case that must pass and one that
%must fail.
%
% res = test_precheck()
%
% The failing cases are the defects from Phase 3A/3B, reproduced on purpose:
% a mask whose length does not match its target (silent in MATLAB), a struct with
% a missing field, a NaN in a numeric result, a product name passed to license().
% A pre-check that only ever says PASS is worse than none.
if nargin < 1, verbose = true; end
here = fileparts(mfilename('fullpath'));
repo = fileparts(fileparts(fileparts(here)));
addpath(fullfile(fileparts(here),'scripts'));
addpath(fullfile(repo,'matlab','boost'));
T = struct('name',{},'kind',{},'args',{},'expect',{},'note',{});
T(end+1) = ck('function present','function',{'boost_params'},true,'');
T(end+1) = ck('function missing','function',{'no_such_fn_xyz'},false,'MISSING_FUNCTION');
s = struct('a',1,'b',2);
T(end+1) = ck('struct complete','struct',{s,{'a','b'}},true,'');
T(end+1) = ck('struct missing field','struct',{s,{'a','c'}},false,'STRUCT_FIELD_MISMATCH');
T(end+1) = ck('numeric finite','numeric',{1:3,struct('finite',true)},true,'');
T(end+1) = ck('numeric NaN','numeric',{[1 NaN 3],struct('finite',true)},false,'NUMERIC_CONSTRAINT');
T(end+1) = ck('numeric range','numeric',{1.5,struct('range',[0 1])},false,'NUMERIC_CONSTRAINT');
% the exact Phase 3B.2 bug: a per-period mask used on a waveform array
T(end+1) = ck('mask length ok','index',{true(3000,1),3000},true,'');
T(end+1) = ck('mask length wrong','index',{true(3000,1),120000},false,'INDEX_MASK_MISMATCH');
T(end+1) = ck('cells uniform','cell',{{1,2,3},[1 1]},true,'');
T(end+1) = ck('cells ragged','cell',{{1,[2 3],4},[]},false,'CELL_SHAPE_MISMATCH');
% product name vs license feature: this product IS licensed on this machine
T(end+1) = ck('toolbox by product name','toolbox',{'Deep Learning Toolbox'},true,'');
% The error class is now UNKNOWN_PRODUCT, from mas_dep_probe. It used to be the
% blanket TOOLBOX_UNAVAILABLE; "the product is not in the table at all" and "the
% product is installed but not licensed" need different fixes, so they are
% different classes. The expected class is updated to the more specific one.
T(end+1) = ck('unknown toolbox','toolbox',{'Quantum Basket Weaver'},false,'UNKNOWN_PRODUCT');
T(end+1) = ck('path exists','path',{fullfile(repo,'matlab','boost')},true,'');
T(end+1) = ck('path missing','path',{fullfile(repo,'no_such_dir')},false,'PATH_MISSING');
mdl = fullfile(repo,'matlab','boost','models','boost_avg.slx');
T(end+1) = ck('model loads','model',{mdl},true,'');
nP = 0;
if verbose, fprintf('=== precheck tests ===\n'); end
R = struct('name',{},'pass',{},'detail',{});
for k = 1:numel(T)
t = T(k);
r = mas_precheck(t.kind, t.args{:});
ok = (r.pass == t.expect);
if ~t.expect && ~isempty(t.note)
ok = ok && strcmp(r.errorClass, t.note);
end
nP = nP + ok;
if verbose
fprintf(' [%s] %-24s %-9s %s\n', tern(ok,'PASS','FAIL'), t.name, ...
t.kind, r.detail);
end
R(end+1) = struct('name',t.name,'pass',ok,'detail',r.detail); %#ok<AGROW>
end
res = struct('nPass',nP,'nTotal',numel(T),'cases',R,'pass',nP==numel(T));
if verbose, fprintf('\n---- %d / %d precheck cases passed ----\n', nP, numel(T)); end
end
% ---------------------------------------------------------------------------
function t = ck(name, kind, args, expect, note)
% struct('args',{args}) would EXPAND the cell into a struct array; assign after
% construction so the cell is stored as one argument list.
t = struct('name',name,'kind',kind,'expect',expect,'note',note);
t.args = args;
end
function s = tern(c,a,b)
if c, s = a; else, s = b; end
end
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |