fix(#2851): replace bare gsd-tools invocations with absolute path (#2869)

* fix(#2851): replace bare gsd-tools invocations with absolute path

`gsd-tools` is not a published bin entry — package.json declares only
get-shit-done-cc and gsd-sdk. The shipped invocation pattern is
`node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" <subcommand>`,
used by every other workflow file.

Two leaked bare invocations:

- get-shit-done/workflows/plan-phase.md §13e (gap-analysis)
  — reported in #2851; gap-analysis silently skipped on every plan-phase run
- get-shit-done/workflows/ingest-docs.md §finalize (commit)
  — caught by the new structural test; ingest-docs commit step was broken

Both updated to canonical absolute-path form.

Adds tests/bug-2851-workflow-bare-gsd-tools.test.cjs which parses every
markdown file under get-shit-done/workflows/, extracts shell-fenced code
blocks, tokenizes each line, and asserts no token in command position is
the bare string `gsd-tools` (the trailing `.cjs` is a different token).
The test also asserts plan-phase.md's gap-analysis call uses the canonical
`node …/gsd-tools.cjs` form.

Closes #2851

* fix(#2851): catch third bare gsd-tools call in ingest-docs.md init

After the first commit, the structural test was strengthened to detect
bare `gsd-tools` inside `$(...)` and backtick command-substitution forms.
The improved test surfaced a third leak:

  ingest-docs.md:55: INIT=$(gsd-tools init ingest-docs)

Fixed to canonical form
  INIT=$(node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" init ingest-docs)
plus the standard `@file:` handoff line that every other workflow uses
when capturing INIT (required by tests/windows-robustness.test.cjs).

Updated tests/bug-2801-ingest-docs-handler.test.cjs to match either
the bare `gsd-tools init ingest-docs` or canonical
`gsd-tools.cjs" init ingest-docs` form — the test's intent is to verify
the dispatch handler is wired, not to lock the bare-bin form that #2851
removes.

Closes #2851

* test(#2851): tighten ingest-docs and gap-analysis assertions to canonical form

CodeRabbit caught two soft assertions in the regression tests:

1. tests/bug-2801: the init-ingest-docs assertion accepted both the
   legacy bare `gsd-tools` form and the canonical node-path form.
   Since #2851 is the fix that removes the bare form, the test should
   only accept the canonical absolute-path invocation. Switched to
   parsed-bash-block extraction with an anchored regex on the full
   `node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs"` path.

2. tests/bug-2851: the gap-analysis assertion used two loose
   .includes()/word-boundary checks. Replaced with a single
   assert.match() against the full canonical path so non-canonical
   forms fail.

* test(#2851): env-assignment skip accepts lowercase identifiers too

CodeRabbit caught: the cmdIdx-skip regex /^[A-Z_][A-Z0-9_]*=/ only
matched uppercase variable names, so a line like `tmp=1 gsd-tools init`
would tokenize to ['tmp=1','gsd-tools','init'], the regex would fail on
'tmp=1', cmdIdx would stay at 0, and the command-position check would
compare 'tmp=1' against 'gsd-tools' — false negative.

POSIX shell variable names are [A-Za-z_][A-Za-z0-9_]*. Widen the regex
to match the actual lexical rule. Existing uppercase forms still work
(FOO=bar gsd-tools); now lowercase forms (tmp=1 gsd-tools) and mixed
case forms are also detected.
This commit is contained in:
Tom Boucher
2026-04-29 21:52:20 -04:00
committed by GitHub
parent ef08a89241
commit 99af76b3ba
4 changed files with 180 additions and 8 deletions

View File

@@ -52,7 +52,8 @@ If `PATH_NOT_FOUND` or `MANIFEST_NOT_FOUND`: display error and exit.
Run the init query:
```bash
INIT=$(gsd-tools init ingest-docs)
INIT=$(node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" init ingest-docs)
if [[ "$INIT" == @file:* ]]; then INIT=$(cat "${INIT#@file:}"); fi
```
Parse `project_exists`, `planning_exists`, `has_git`, `project_path` from INIT.
@@ -289,7 +290,8 @@ Preview the merge diff to the user and gate via approve-revise-abort before writ
Commit the ingest results:
```bash
gsd-tools commit "docs: ingest {N} docs from {SCAN_PATH} (#2387)" --files \
node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" commit \
"docs: ingest {N} docs from {SCAN_PATH} (#2387)" --files \
.planning/PROJECT.md \
.planning/REQUIREMENTS.md \
.planning/ROADMAP.md \

View File

@@ -1438,11 +1438,11 @@ one place before execution begins.
```bash
POST_PLANNING_GAPS=$(gsd-sdk query config-get workflow.post_planning_gaps --default true 2>/dev/null || echo true)
if [ "$POST_PLANNING_GAPS" = "true" ]; then
gsd-tools gap-analysis --phase-dir "${PHASE_DIR}"
node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" gap-analysis --phase-dir "${PHASE_DIR}"
fi
```
(`gsd-tools gap-analysis` reads `.planning/REQUIREMENTS.md`, `${PHASE_DIR}/CONTEXT.md`,
(`gsd-tools.cjs gap-analysis` reads `.planning/REQUIREMENTS.md`, `${PHASE_DIR}/CONTEXT.md`,
and `${PHASE_DIR}/*-PLAN.md`, then prints a markdown table with one row per
REQ-ID and D-ID. Word-boundary matching prevents `REQ-1` from being mistaken for
`REQ-10`.)

View File

@@ -126,11 +126,19 @@ describe('bug-2801: ingest-docs.md workflow calls gsd-tools not gsd-sdk', () =>
);
});
test('ingest-docs.md init step uses gsd-tools init ingest-docs', () => {
test('ingest-docs.md init step uses canonical node-path gsd-tools.cjs invocation', () => {
const content = fs.readFileSync(WORKFLOW_FILE, 'utf-8');
const lines = content.split('\n');
const initLine = lines.find(l => /gsd-tools\s+init\s+ingest-docs/.test(l));
assert.ok(initLine, 'workflow must contain "gsd-tools init ingest-docs"');
// Parse fenced bash blocks structurally — do not match raw markdown text.
const codeBlockRe = /```bash\n([\s\S]*?)```/g;
const bashLines = [...content.matchAll(codeBlockRe)]
.flatMap((m) => m[1].split('\n'))
.filter((l) => !/^\s*#/.test(l));
// Per #2851 the only valid form is the absolute-path node invocation; the
// legacy bare `gsd-tools` is the bug being fixed and must not be accepted.
const initLine = bashLines.find((l) =>
/\bnode\s+["']?\$HOME\/\.claude\/get-shit-done\/bin\/gsd-tools\.cjs["']?\s+init\s+ingest-docs\b/.test(l)
);
assert.ok(initLine, 'workflow must invoke init ingest-docs via canonical node-path gsd-tools.cjs');
});
test('cmdInitIngestDocs is exported from init.cjs', () => {

View File

@@ -0,0 +1,162 @@
/**
* Bug #2851: plan-phase.md §13e calls bare `gsd-tools` — incomplete fix of #2245
*
* `gsd-tools` is NOT a published bin entry. The shipped invocation pattern is:
*
* node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" <subcommand> [args]
*
* Some workflow markdown files leaked the bare `gsd-tools <subcommand>` form,
* which fails with `command not found` at runtime.
*
* This test parses every markdown file in get-shit-done/workflows/ structurally:
* it tokenizes the content into fenced code blocks, then on each shell-block
* line checks whether `gsd-tools` appears as a bare command (not preceded by
* `node `, not part of the filename `gsd-tools.cjs`, not inside a comment).
*
* Per project rule: this test does NOT use grep/regex .includes() on raw file
* content as the assertion surface. Instead, it splits into code-fenced blocks
* and tokenizes each line — only command-position tokens count as violations.
*/
'use strict';
const { describe, test } = require('node:test');
const assert = require('node:assert/strict');
const fs = require('node:fs');
const path = require('node:path');
const WORKFLOWS_DIR = path.join(__dirname, '..', 'get-shit-done', 'workflows');
/**
* Extract shell-fenced code blocks from a markdown file.
* Returns an array of { startLine, lines } where lines are the contents
* between the ```bash / ```sh / ```shell fence markers.
*/
function extractShellBlocks(content) {
const allLines = content.split('\n');
const blocks = [];
let inBlock = false;
let blockLang = null;
let blockStart = 0;
let blockLines = [];
for (let i = 0; i < allLines.length; i++) {
const line = allLines[i];
const fenceOpen = line.match(/^```(\w+)?/);
if (!inBlock && fenceOpen) {
inBlock = true;
blockLang = (fenceOpen[1] || '').toLowerCase();
blockStart = i + 2; // 1-indexed line number of first content line
blockLines = [];
continue;
}
if (inBlock && /^```\s*$/.test(line)) {
if (['bash', 'sh', 'shell', 'zsh', ''].includes(blockLang)) {
blocks.push({ startLine: blockStart, lines: blockLines });
}
inBlock = false;
blockLang = null;
blockLines = [];
continue;
}
if (inBlock) {
blockLines.push(line);
}
}
return blocks;
}
/**
* Check a single shell-block line for a bare `gsd-tools` command-position token.
* Returns true if the line is a violation.
*/
function lineHasBareGsdTools(line) {
// Strip leading whitespace and any prompt prefix ($ , > , # )
let l = line.replace(/^\s*[$>]\s*/, '');
// Skip pure comment lines
if (/^\s*#/.test(l)) return false;
// Strip inline comment (# preceded by whitespace, not inside a string)
// Conservative: only strip if # appears after whitespace and outside quotes —
// we just look for the first ` #` outside of quoted context. For our needs,
// splitting on `^[^"']*?(\s#)` is good enough.
const hashIdx = l.search(/(?:^|[^"'\w])#/);
if (hashIdx > 0) l = l.slice(0, hashIdx);
// Unwrap command-substitution forms so the substituted command is in
// command position. `$(cmd …)` and `` `cmd …` `` both run the inner string
// as a fresh command, so a bare `gsd-tools` inside them is just as broken
// as one at the start of the line. Iterate until stable for nested forms.
let prev;
do {
prev = l;
l = l.replace(/\$\(([^()]*)\)/g, ' $1 ').replace(/`([^`]*)`/g, ' $1 ');
} while (l !== prev);
// Tokenize on whitespace, semicolons, pipes, and && / ||
// Then walk tokens — a violation is a token that starts with `gsd-tools`
// followed by a word boundary (so `gsd-tools.cjs` does NOT match), and the
// preceding token is NOT `node`.
const segments = l.split(/(?:\s*(?:&&|\|\||;|\|)\s*)/);
for (const seg of segments) {
const tokens = seg.trim().split(/\s+/).filter(Boolean);
if (tokens.length === 0) continue;
// Skip env var assignments at the start (FOO=bar gsd-tools …, tmp=1 gsd-tools …).
// POSIX shell variable names are [A-Za-z_][A-Za-z0-9_]*; lowercase is valid.
let cmdIdx = 0;
while (cmdIdx < tokens.length && /^[A-Za-z_][A-Za-z0-9_]*=/.test(tokens[cmdIdx])) {
cmdIdx++;
}
if (cmdIdx >= tokens.length) continue;
const cmd = tokens[cmdIdx];
// Match `gsd-tools` exactly (no extension), as command position.
if (cmd === 'gsd-tools') return true;
}
return false;
}
describe('bug-2851: workflow files must not call bare `gsd-tools` (#2245 sweep regression)', () => {
test('no get-shit-done/workflows/*.md file contains a bare gsd-tools command', () => {
const files = fs.readdirSync(WORKFLOWS_DIR).filter((f) => f.endsWith('.md'));
assert.ok(files.length > 0, 'expected workflow files to exist');
const violations = [];
for (const f of files) {
const full = path.join(WORKFLOWS_DIR, f);
const content = fs.readFileSync(full, 'utf-8');
const blocks = extractShellBlocks(content);
for (const blk of blocks) {
for (let i = 0; i < blk.lines.length; i++) {
if (lineHasBareGsdTools(blk.lines[i])) {
violations.push(`${f}:${blk.startLine + i}: ${blk.lines[i].trim()}`);
}
}
}
}
assert.deepStrictEqual(
violations,
[],
'Bare `gsd-tools` invocations found in workflow shell blocks. ' +
'Use `node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" <subcommand>` instead.\n' +
violations.join('\n'),
);
});
test('plan-phase.md §13e gap-analysis uses canonical absolute-path invocation', () => {
const planPhase = fs.readFileSync(path.join(WORKFLOWS_DIR, 'plan-phase.md'), 'utf-8');
const blocks = extractShellBlocks(planPhase);
let foundGapAnalysisCall = false;
for (const blk of blocks) {
for (const line of blk.lines) {
if (/gap-analysis/.test(line) && !/^\s*#/.test(line)) {
foundGapAnalysisCall = true;
assert.match(
line,
/\bnode\s+["']?\$HOME\/\.claude\/get-shit-done\/bin\/gsd-tools\.cjs["']?\s+gap-analysis\b/,
`gap-analysis call must use canonical absolute-path invocation, got: ${line.trim()}`,
);
}
}
}
assert.ok(foundGapAnalysisCall, 'expected at least one gap-analysis invocation in plan-phase.md');
});
});