fix(#2559): remove stale year injection from research agent web search instructions (#2591)

The gsd-phase-researcher and gsd-project-researcher agents instructed
WebSearch queries to always include 'current year' (e.g., 2024). As
time passes, a hardcoded year biases search results toward stale
dated content — users saw 2024-tagged queries producing stale blog
references in 2026.

Remove the year-injection guidance. Instead, rely on checking
publication dates on the returned sources. Query templates and
success criteria updated accordingly.

Closes #2559
This commit is contained in:
Tom Boucher
2026-04-22 12:04:13 -04:00
committed by GitHub
parent 67a9550720
commit 6b7b5c15a5
3 changed files with 78 additions and 5 deletions

View File

@@ -145,7 +145,7 @@ When researching "best library for X": find what the ecosystem actually uses, do
1. `mcp__context7__resolve-library-id` with libraryName
2. `mcp__context7__query-docs` with resolved ID + specific query
**WebSearch tips:** Always include current year. Use multiple query variations. Cross-verify with authoritative sources.
**WebSearch tips:** Use multiple query variations. Cross-verify with authoritative sources. Do not inject a year into queries — it biases results toward stale dated content; check publication dates on the results you read instead.
## Enhanced Web Search (Brave API)
@@ -836,6 +836,6 @@ Quality indicators:
- **Verified, not assumed:** Findings cite Context7 or official docs
- **Honest about gaps:** LOW confidence items flagged, unknowns admitted
- **Actionable:** Planner could create tasks based on this research
- **Current:** Year included in searches, publication dates checked
- **Current:** Publication dates checked on sources (do not inject year into queries)
</success_criteria>

View File

@@ -116,12 +116,12 @@ For finding what exists, community patterns, real-world usage.
**Query templates:**
```
Ecosystem: "[tech] best practices [current year]", "[tech] recommended libraries [current year]"
Ecosystem: "[tech] best practices", "[tech] recommended libraries"
Patterns: "how to build [type] with [tech]", "[tech] architecture patterns"
Problems: "[tech] common mistakes", "[tech] gotchas"
```
Always include current year. Use multiple query variations. Mark WebSearch-only findings as LOW confidence.
Use multiple query variations. Mark WebSearch-only findings as LOW confidence. Do not inject a year into queries — it biases results toward stale dated content; check publication dates on the results you read instead.
### Enhanced Web Search (Brave API)
@@ -672,6 +672,6 @@ Research is complete when:
- [ ] Files written (DO NOT commit — orchestrator handles this)
- [ ] Structured return provided to orchestrator
**Quality:** Comprehensive not shallow. Opinionated not wishy-washy. Verified not assumed. Honest about gaps. Actionable for roadmap. Current (year in searches).
**Quality:** Comprehensive not shallow. Opinionated not wishy-washy. Verified not assumed. Honest about gaps. Actionable for roadmap. Current (check publication dates, do not inject year into queries).
</success_criteria>

View File

@@ -0,0 +1,73 @@
/**
* Bug #2559: Stale document references in Research phase
*
* The gsd-phase-researcher and gsd-project-researcher agents instruct
* WebSearch queries to always include "current year" (or a hardcoded
* year). This biases results toward stale dated content as time passes
* (e.g., a 2024 query run in 2026 returns stale results).
*
* Fix: Remove year-injection instructions from research agent
* WebSearch guidance so searches return current results.
*/
const { describe, test } = require('node:test');
const assert = require('node:assert/strict');
const fs = require('fs');
const path = require('path');
const PHASE_RESEARCHER = path.join(
__dirname,
'..',
'agents',
'gsd-phase-researcher.md'
);
const PROJECT_RESEARCHER = path.join(
__dirname,
'..',
'agents',
'gsd-project-researcher.md'
);
const FILES = [
{ label: 'gsd-phase-researcher.md', path: PHASE_RESEARCHER },
{ label: 'gsd-project-researcher.md', path: PROJECT_RESEARCHER },
];
describe('research agents do not inject year into web searches (#2559)', () => {
for (const { label, path: filePath } of FILES) {
test(`${label} contains no CURRENT_YEAR placeholder`, () => {
const content = fs.readFileSync(filePath, 'utf-8');
assert.ok(
!/CURRENT_YEAR/.test(content),
`${label} must not contain CURRENT_YEAR placeholder (causes stale-year injection)`
);
});
test(`${label} contains no hardcoded year in web search instructions`, () => {
const content = fs.readFileSync(filePath, 'utf-8');
const match = content.match(/\b20(2[3-9]|[3-9]\d)\b/);
assert.ok(
!match,
`${label} must not contain hardcoded year (found "${match && match[0]}") — biases searches toward stale content`
);
});
test(`${label} does not instruct searches to include year or current year`, () => {
const content = fs.readFileSync(filePath, 'utf-8');
// Match phrases like "include current year", "year in searches",
// "[current year]", "with year", etc.
const patterns = [
/include\s+(?:the\s+)?current\s+year/i,
/current\s+year/i,
/year\s+in\s+(?:searches|queries)/i,
/\[current year\]/i,
];
for (const pat of patterns) {
assert.ok(
!pat.test(content),
`${label} must not instruct year injection (matched /${pat.source}/)`
);
}
});
}
});