@aabadin/project-memory-context 0.2.1 → 0.2.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/package.json
CHANGED
|
@@ -108,6 +108,7 @@ async function installOpencode({ projectRoot, packageRoot, placeholders, globalC
|
|
|
108
108
|
'opencode/commands/enrich-status.md',
|
|
109
109
|
'opencode/commands/doctor.md',
|
|
110
110
|
'opencode/commands/init-project.md',
|
|
111
|
+
'opencode/commands/retry-errors.md',
|
|
111
112
|
];
|
|
112
113
|
|
|
113
114
|
for (const tpl of commandTemplates) {
|
|
@@ -0,0 +1,31 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: retry-errors
|
|
3
|
+
description: Re-enrich symbols that previously failed using Ollama directly, with a detailed per-symbol report.
|
|
4
|
+
argument-hint: "[--limit N] [--model MODEL] [--concurrency N] [--timeout MS]"
|
|
5
|
+
allowed-tools:
|
|
6
|
+
- Bash
|
|
7
|
+
---
|
|
8
|
+
|
|
9
|
+
<objective>
|
|
10
|
+
Retry enrichment for all symbols currently in error status, using only the local Ollama provider. Generates a per-symbol report showing what each previous error was and whether the retry succeeded.
|
|
11
|
+
</objective>
|
|
12
|
+
|
|
13
|
+
<execution>
|
|
14
|
+
Run:
|
|
15
|
+
|
|
16
|
+
```bash
|
|
17
|
+
{{PMC_BIN}} retry-errors . --model "qwen2.5-coder:14b" --concurrency 1 --timeout 300000
|
|
18
|
+
```
|
|
19
|
+
|
|
20
|
+
Options:
|
|
21
|
+
- `--limit N` — Only retry the first N error entries (useful for testing)
|
|
22
|
+
- `--model MODEL` — Ollama model to use (default: from config)
|
|
23
|
+
- `--concurrency N` — Parallel slots (recommended: 1 for Ollama)
|
|
24
|
+
- `--timeout MS` — Timeout per symbol in ms (default: 600000 = 10min)
|
|
25
|
+
- `--report PATH` — Custom path for the JSON report
|
|
26
|
+
|
|
27
|
+
The command reads the worklist, finds all entries with `status: "error"`, and re-enriches them using only the local-model (Ollama) provider. A JSON report is saved to `.planning/project-memory-context/enrichment/retry-report.json` with:
|
|
28
|
+
- Previous error details for each symbol (error type, message, provider)
|
|
29
|
+
- Retry result (succeeded/failed, elapsed time, content preview)
|
|
30
|
+
- Overall summary counts
|
|
31
|
+
</execution>
|