result-bitable-reporter

Original：🇺🇸 English

Translated

1 scripts

Collect app events via evalpkgs into sqlite, then filter/report capture_results to Feishu Bitable with retry-safe writeback. Use for collect-start/collect-stop/filter/report/retry-reset workflows.

1installs

Sourcehttprunner/skills

Added on2026-02-08

NPX Install

npx skill4agent add httprunner/skills result-bitable-reporter

SKILL.md Content

View Translation Comparison →

Result Bitable Reporter

Use this skill for a deterministic

sqlite -> Feishu

result pipeline around

capture_results

.

Path Convention

Canonical install and execution directory:

~/.agents/skills/result-bitable-reporter/

. Run commands from this directory:

bash

cd ~/.agents/skills/result-bitable-reporter

One-off (safe in scripts/loops from any working directory):

bash

(cd ~/.agents/skills/result-bitable-reporter && npx tsx scripts/result_reporter.ts --help)

Workflow

Optional data collection:

```
collect-start
```
: start background
```
evalpkgs run
```
for one device (
```
SerialNumber
```
) with
```
TaskID
```
(digits only).

collect-stop

: stop that collector and print summary metrics (

delta

,

task_delta

,

records_jsonl

,

tracking_events

,

runtime_sec

).

Data selection:

```
filter
```
: preview target sqlite rows (
```
capture_results
```
) before upload.
Task-scoped selection: pass
```
--task-id <TASK_ID>
```
to constrain rows to one task.
Default status filter is pending+failed (
```
reported IN (0, -1)
```
).

Data reporting:

```
report
```
: batch create Feishu records and write back sqlite status.
Optional
```
--max-rows <n>
```
sets total cap for one report run (default: process all matched rows).
For per-task workflows, always pass
```
--task-id <TASK_ID>
```
to avoid cross-task uploads.

Success writeback:

reported=1

,

reported_at=now_ms

,

report_error=NULL

.

Failure writeback:

reported=-1

,

reported_at=now_ms

,

report_error=<truncated error>

.

Retry:

```
retry-reset
```
: move failed rows (
```
reported=-1
```
) back to pending (
```
reported=0
```
), then rerun
```
report
```
.

Run

CLI entry:

bash

npx tsx scripts/result_reporter.ts <subcommand> [flags]

Subcommands:

```
collect-start
```
```
collect-stop
```
```
filter
```
```
report
```
```
retry-reset
```

Required env by phase:

Collection:
```
BUNDLE_ID
```
,
```
SerialNumber
```

Feishu report:

FEISHU_APP_ID

,

FEISHU_APP_SECRET

,

RESULT_BITABLE_URL

Optional overrides:

FEISHU_BASE_URL

,

TRACKING_STORAGE_DB_PATH

,

RESULT_SQLITE_TABLE

Resources

```
scripts/result_reporter.ts
```
: executable source of truth for flags and behavior.
```
references/sqlite-and-field-mapping.md
```
: sqlite schema, collect-start/collect-stop semantics, SQL operations, command examples.
```
references/feishu-api-and-errors.md
```
: Feishu APIs, URL rules, failures, field mapping env overrides, command examples.

result-bitable-reporter

NPX Install

Tags

SKILL.md Content

Result Bitable Reporter

Path Convention

Workflow

Run

Resources