EPA EIS Records Fetch
Use the CLI-owned epa.eis-records/search operation. This Skill supplies intent
routing, search-selection guidance, and result-use boundaries only; the CLI
TypeScript 7 runtime owns official endpoint validation, HTML parsing, schemas,
limits, retries, deduplication, partial results, and receipts.
Before running
- Read
references/tiangong-data-requirement.json.
- Use the caller- or workspace-resolved stable CLI. The requirement declares
compatible capability and operation contract majors; it does not select a
package build.
- Run
data describe with that same CLI. Continue only when the capability
ID and required contract majors match, and copy the exact current
capability/operation versions from that response into the run request.
tiangong-ai data describe epa.eis-records --json
Use the returned Discovery Metadata to confirm current source coverage,
granularity, HTML/provider limitations, selection hints, provides, and
doesNotProvide. Do not rely on objective source facts copied from an older
Skill revision.
Choose the search
- Use
openComment for the provider's current open-comment queue,
last30Published for recently published records, last60Issued for recently
issued records, or lastWeek for the provider's last-week view.
- For a precise project or historical query, first build and review that search
in the official EPA EIS Database UI, then copy its complete HTTPS search URL
into
searchUrls. The CLI rejects non-EPA origins and non-search paths.
- Common searches run before explicit URLs, while each group preserves caller
order. Use more than one only when the evidence need and coverage plan call
for multiple provider result surfaces.
- Empty or sparse output applies only to the selected search page. It is not
proof that EPA, CEQ, another agency, or the real world has no relevant EIS.
Prepare the request
Build one tiangong.data.run-request.v1 envelope. Replace the version
placeholders with the exact versions from the same data describe response and validate input against the current
schema returned by data describe. The example demonstrates both selector
types; remove the one not required for the task.
{
"schemaVersion": "tiangong.data.run-request.v1",
"capabilityId": "epa.eis-records",
"capabilityVersion": "<describe.manifest.capabilityVersion>",
"operationId": "search",
"operationVersion": "<describe.manifest.operations[0].operationVersion>",
"input": {
"commonSearches": ["openComment"],
"searchUrls": [
"https://cdxapps.epa.gov/cdx-enepa-II/public/action/eis/search?search=Glen+Canyon&state=CO"
]
}
}
Use run-request limits only when intentionally lowering the manifest's
search-page or record ceiling. Do not place output paths, retry settings, a
custom base URL, credentials, or document download instructions in input.
Run
tiangong-ai data run epa.eis-records search \
--input /absolute/path/to/request.json --json
Preserve the complete tiangong.data.run-result.v1 envelope, including
contract, warnings, errors, observations, and receipt, with data.
Result boundaries
- Treat
partial as a failed later search surface: retain successful records
and report the missing search. Treat blocked as no usable result, including
an unrecognized provider page after HTML drift.
- Preserve page provenance, CEQ number, unique ID, document type, dates,
agencies, state, detail URL, and download cues. Deduplication is by official
identifiers; it does not establish that similarly titled records are the
same project or action.
- Download links and IDs are availability cues only. Do not claim any linked
EIS, notice, comment letter, or attachment was downloaded, parsed, reviewed,
or admitted as evidence.
- Provider metadata does not decide NEPA adequacy, legal sufficiency,
environmental effects, agency compliance, causality, policy responsibility,
or report conclusions. Those require separately governed document
acquisition and analysis.
- HTML parsing, durable artifact placement, document acquisition,
normalization, evidence admission, and cross-source synthesis remain outside
this thin Skill.
Reference
references/tiangong-data-requirement.json: stable capability requirement; it is not a package lock.
1---2name: epa-eis-records-fetch3description: Retrieve official U.S. EPA EIS Database result-table metadata through the Tiangong CLI. Use when a task needs Environmental Impact Statement titles, CEQ/provider IDs, document type, dates, agencies, state, detail links, or document-availability cues from a current common search or a precise search URL created in the official UI; do not use to claim linked documents were reviewed or to judge NEPA adequacy, legal sufficiency, environmental effects, compliance, or policy responsibility.4---5
6# EPA EIS Records Fetch
7
8Use the CLI-owned `epa.eis-records/search` operation. This Skill supplies intent
9routing, search-selection guidance, and result-use boundaries only; the CLI
10TypeScript 7 runtime owns official endpoint validation, HTML parsing, schemas,
11limits, retries, deduplication, partial results, and receipts.
12
13## Before running
14
151. Read `references/tiangong-data-requirement.json`.
162. Use the caller- or workspace-resolved stable CLI. The requirement declares
17 compatible capability and operation contract majors; it does not select a
18 package build.
193. Run `data describe` with that same CLI. Continue only when the capability
20 ID and required contract majors match, and copy the exact current
21 capability/operation versions from that response into the run request.
22
23```bash
24tiangong-ai data describe epa.eis-records --json
25```
26
27Use the returned Discovery Metadata to confirm current source coverage,
28granularity, HTML/provider limitations, selection hints, `provides`, and
29`doesNotProvide`. Do not rely on objective source facts copied from an older
30Skill revision.
31
32## Choose the search
33
34- Use `openComment` for the provider's current open-comment queue,
35 `last30Published` for recently published records, `last60Issued` for recently
36 issued records, or `lastWeek` for the provider's last-week view.
37- For a precise project or historical query, first build and review that search
38 in the official EPA EIS Database UI, then copy its complete HTTPS search URL
39 into `searchUrls`. The CLI rejects non-EPA origins and non-search paths.
40- Common searches run before explicit URLs, while each group preserves caller
41 order. Use more than one only when the evidence need and coverage plan call
42 for multiple provider result surfaces.
43- Empty or sparse output applies only to the selected search page. It is not
44 proof that EPA, CEQ, another agency, or the real world has no relevant EIS.
45
46## Prepare the request
47
48Build one `tiangong.data.run-request.v1` envelope. Replace the version
49placeholders with the exact versions from the same `data describe` response and validate `input` against the current
50schema returned by `data describe`. The example demonstrates both selector
51types; remove the one not required for the task.
52
53```json
54{
55 "schemaVersion": "tiangong.data.run-request.v1",
56 "capabilityId": "epa.eis-records",
57 "capabilityVersion": "<describe.manifest.capabilityVersion>",
58 "operationId": "search",
59 "operationVersion": "<describe.manifest.operations[0].operationVersion>",
60 "input": {
61 "commonSearches": ["openComment"],
62 "searchUrls": [
63 "https://cdxapps.epa.gov/cdx-enepa-II/public/action/eis/search?search=Glen+Canyon&state=CO"
64 ]
65 }
66}
67```
68
69Use run-request `limits` only when intentionally lowering the manifest's
70search-page or record ceiling. Do not place output paths, retry settings, a
71custom base URL, credentials, or document download instructions in `input`.
72
73## Run
74
75```bash
76tiangong-ai data run epa.eis-records search \
77 --input /absolute/path/to/request.json --json
78```
79
80Preserve the complete `tiangong.data.run-result.v1` envelope, including
81`contract`, `warnings`, `errors`, `observations`, and `receipt`, with `data`.
82
83## Result boundaries
84
85- Treat `partial` as a failed later search surface: retain successful records
86 and report the missing search. Treat `blocked` as no usable result, including
87 an unrecognized provider page after HTML drift.
88- Preserve page provenance, CEQ number, unique ID, document type, dates,
89 agencies, state, detail URL, and download cues. Deduplication is by official
90 identifiers; it does not establish that similarly titled records are the
91 same project or action.
92- Download links and IDs are availability cues only. Do not claim any linked
93 EIS, notice, comment letter, or attachment was downloaded, parsed, reviewed,
94 or admitted as evidence.
95- Provider metadata does not decide NEPA adequacy, legal sufficiency,
96 environmental effects, agency compliance, causality, policy responsibility,
97 or report conclusions. Those require separately governed document
98 acquisition and analysis.
99- HTML parsing, durable artifact placement, document acquisition,
100 normalization, evidence admission, and cross-source synthesis remain outside
101 this thin Skill.
102
103## Reference
104
105- `references/tiangong-data-requirement.json`: stable capability requirement; it is not a package lock.