AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

database-lookup

Queries documented public database APIs with explicit endpoints, filters, pagination, and provenance. Used when a scientific, regulatory, financial, or other database-backed fact must be retrieved reproducibly from a named source rather than inferred from general knowledge.

Original source
K-Dense-AI
License
MIT
Pinned revision
5f2bc0341354fd5590aef76a533ada1e1e5b4e62

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

Ask your assistant

Use Agentic Memory skills to read k-dense-ai--database-lookup at revision 5f2bc0341354fd5590aef76a533ada1e1e5b4e62, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 83

references/simbad.md

SIMBAD Astronomical Database (CDS Strasbourg)

SIMBAD contains data on astronomical objects beyond the Solar System, including identifications, coordinates, photometry, proper motions, parallaxes, radial velocities, spectral types, and bibliographic references.

Base URLs

TAP endpoint (recommended):

https://simbad.cds.unistra.fr/simbad/sim-tap/sync

Script interface:

https://simbad.cds.unistra.fr/simbad/sim-script

Simple query endpoints:

https://simbad.cds.unistra.fr/simbad/sim-id
https://simbad.cds.unistra.fr/simbad/sim-coo

Authentication

No API key required. All endpoints are public.

Key Endpoints

1. TAP Queries (ADQL — recommended for programmatic use)

POST /simbad/sim-tap/sync
Content-Type: application/x-www-form-urlencoded

Parameters:
  REQUEST=doQuery
  LANG=ADQL
  QUERY=<adql query>
  FORMAT=json|votable|csv|tsv
  MAXREC=<max rows>

| Parameter | Type | Description | |-----------|--------|-------------| | REQUEST | string | doQuery | | LANG | string | ADQL | | FORMAT | string | json, votable, csv, tsv. | | QUERY | string | Required. ADQL query. | | MAXREC | int | Max rows returned. Always set to avoid downloading millions of rows. |

Example — look up object by name:

https://simbad.cds.unistra.fr/simbad/sim-tap/sync?request=doQuery&lang=adql&format=json&query=SELECT basic.OID, ra, dec, main_id, otype FROM basic JOIN ident ON oid = ident.oidref WHERE id = 'M31'

Example — cone search (objects within 5 arcmin of coordinates):

https://simbad.cds.unistra.fr/simbad/sim-tap/sync?request=doQuery&lang=adql&format=json&query=SELECT TOP 50 main_id, ra, dec, otype FROM basic WHERE CONTAINS(POINT('ICRS', ra, dec), CIRCLE('ICRS', 10.684, 41.269, 0.083)) = 1

Note: radius in CIRCLE is in degrees (5 arcmin = 0.083 deg).

Example — objects by type (e.g., all pulsars):

https://simbad.cds.unistra.fr/simbad/sim-tap/sync?request=doQuery&lang=adql&format=json&query=SELECT TOP 100 main_id, ra, dec, otype FROM basic WHERE otype = 'Psr'

2. Identifier Query (simple lookup)

GET /simbad/sim-id?Ident={name}&output.format=votable

| Parameter | Type | Description | |------------------|--------|-------------| | Ident | string | Required. Object name (e.g., M31, Sirius, NGC 1275). | | output.format | string | votable, html. |

3. Coordinate Query

GET /simbad/sim-coo?Coord={coords}&Radius={radius}&Radius.unit={unit}&output.format=votable

| Parameter | Type | Description | |----------------|--------|-------------| | Coord | string | Required. Coordinates, e.g., 10.684 +41.269 or 00 42 44 +41 16 09. | | Radius | float | Search radius. Default: 2. | | Radius.unit | string | arcmin, arcsec, deg. Default: arcmin. | | output.format| string | votable, html. |

4. Script Interface (for multi-command queries)

POST /simbad/sim-script
Content-Type: application/x-www-form-urlencoded

Body: script=<script text>

A script consists of configuration lines followed by query commands:

output console=off script=off
format object "<format string>"
query id <object name>

Query commands:

  • query id <name> — lookup by name (e.g., query id M31)
  • query coo <ra> <dec> radius=<value><unit> — cone search (units: d=deg, m=arcmin, s=arcsec)
  • query id wildcard <pattern> — wildcard search (e.g., query id wildcard NGC 10*)
  • query sample <criteria> — criteria search (e.g., query sample otype='*' & Vmag < 5.0)

Multi-object queries — include multiple query id lines in a single script:

output console=off script=off
format object "%IDLIST(1) | %COO(s;A D;ICRS) | %OTYPE"
query id M31
query id M42
query id M101

Script Format Codes

Format codes define which fields appear in script output. Use inside format object "...".

Identification

| Code | Description | Example Output | |------|-------------|----------------| | %IDLIST(1) | Primary identifier | M 31 | | %IDLIST | All identifiers | M 31, NGC 224, UGC 454, ... | | %MAIN_ID | Main identifier | M 31 |

Coordinates

| Code | Description | Example Output | |------|-------------|----------------| | %COO(s;A D;ICRS) | RA Dec ICRS (sexagesimal) | 00 42 44.330 +41 16 07.50 | | %COO(d;A D;ICRS) | RA Dec decimal degrees | 10.6847083 +41.2687500 | | %COO(d;A D;GAL) | Galactic coordinates | 121.1743 -21.5733 |

Object Properties

| Code | Description | Example Output | |------|-------------|----------------| | %OTYPE | Object type (short code) | G | | %SP | Spectral type | A1V | | %MT | Morphological type | SA(s)b |

Photometry

| Code | Description | |------|-------------| | %FLUXLIST(V) | V-band magnitude | | %FLUXLIST(B) | B-band magnitude | | %FLUXLIST(U,B,V,R,I) | Multiple bands | | %FLUXLIST(J,H,K) | Near-infrared bands |

Kinematics

| Code | Description | |------|-------------| | %PM | Proper motion (mas/yr) | | %PLX | Parallax (mas) | | %RV | Radial velocity (km/s) |

Predefined Format Levels

Basic:    "%IDLIST(1) | %COO(s;A D;ICRS) | %OTYPE"
Detailed: "%IDLIST(1) | %COO(s;A D;ICRS) | %OTYPE | %SP | %FLUXLIST(V)"
Full:     "%IDLIST(1) | %COO(s;A D;ICRS;J2000) | %OTYPE | %SP | %FLUXLIST(U,B,V,R,I,J,H,K) | %PM | %PLX | %RV | %MT"

Key TAP Tables

basic — Main Object Table

| Column | Type | Description | |--------|------|-------------| | oid | BIGINT | Internal object identifier (primary key) | | main_id | VARCHAR | Primary object identifier | | ra | DOUBLE | Right Ascension in degrees (ICRS) | | dec | DOUBLE | Declination in degrees (ICRS) | | otype | VARCHAR | Condensed object type code | | sp_type | VARCHAR | Spectral type | | plx_value | DOUBLE | Parallax in milliarcseconds | | plx_err | DOUBLE | Parallax error | | pmra | DOUBLE | Proper motion in RA (mas/yr) | | pmdec | DOUBLE | Proper motion in Dec (mas/yr) | | rvz_radvel | DOUBLE | Radial velocity (km/s) | | rvz_err | DOUBLE | Radial velocity error | | galdim_majaxis | DOUBLE | Galaxy major axis (arcmin) | | galdim_minaxis | DOUBLE | Galaxy minor axis (arcmin) | | galdim_angle | DOUBLE | Galaxy position angle (degrees) |

ident — Identifier Table

| Column | Type | Description | |--------|------|-------------| | oidref | BIGINT | Reference to basic.oid | | id | VARCHAR | Identifier string |

flux — Photometric Measurements

| Column | Type | Description | |--------|------|-------------| | oidref | BIGINT | Reference to basic.oid | | filter | VARCHAR | Filter name (U, B, V, R, I, J, H, K, u, g, r, i, z, G, etc.) | | flux | DOUBLE | Magnitude value | | flux_err | DOUBLE | Magnitude error | | bibcode | VARCHAR | Source reference bibcode |

mesDistance — Distance Measurements

| Column | Type | Description | |--------|------|-------------| | oidref | BIGINT | Reference to basic.oid | | dist | DOUBLE | Distance value | | unit | VARCHAR | Distance unit (pc, kpc, Mpc) | | minus_err | DOUBLE | Lower error | | plus_err | DOUBLE | Upper error | | method | VARCHAR | Measurement method | | bibcode | VARCHAR | Source reference |

has_ref / ref — Bibliographic References

has_ref links objects to references (oidref → basic.oid, oidbibref → ref.oidbib).

| ref Column | Type | Description | |--------|------|-------------| | oidbib | BIGINT | Bibliography object ID | | bibcode | VARCHAR | ADS bibcode | | title | VARCHAR | Paper title | | journal | VARCHAR | Journal name | | "year" | INTEGER | Publication year |

otypedef — Object Type Definitions

| Column | Type | Description | |--------|------|-------------| | otype | VARCHAR | Object type code | | description | VARCHAR | Human-readable description |

Common Object Types (otype)

| Code | Description | |------|-------------| | * | Star | | HII | HII region | | PN | Planetary nebula | | SNR | Supernova remnant | | G | Galaxy | | AGN | Active galactic nucleus | | QSO | Quasar | | ClG | Galaxy cluster | | GlC | Globular cluster | | OpC | Open cluster | | Psr | Pulsar | | WD* | White dwarf | | Pl | Extra-solar planet | | ** | Double/multiple star | | V* | Variable star | | X | X-ray source |

Query SELECT * FROM otypedef ORDER BY otype for the full list. otype is the stored short code; label contains display labels such as Galaxy. Equality to * selects only that exact type, not every stellar subtype; use the type hierarchy deliberately when including descendants.

ADQL Query Patterns

Spatial Queries

Cone search — objects within a radius of a point:

SELECT main_id, ra, dec, otype
FROM basic
WHERE CONTAINS(POINT('ICRS', ra, dec), CIRCLE('ICRS', 83.633, 22.014, 0.5)) = 1

Parameters: CIRCLE('ICRS', center_ra_deg, center_dec_deg, radius_deg)

Box search:

SELECT main_id, ra, dec, otype
FROM basic
WHERE CONTAINS(POINT('ICRS', ra, dec), BOX('ICRS', 180.0, 0.0, 10.0, 5.0)) = 1

Parameters: BOX('ICRS', center_ra, center_dec, width_deg, height_deg)

Polygon search:

SELECT main_id, ra, dec
FROM basic
WHERE CONTAINS(POINT('ICRS', ra, dec), POLYGON('ICRS', 10.0, 40.0, 12.0, 40.0, 12.0, 42.0, 10.0, 42.0)) = 1

Angular distance:

SELECT main_id, ra, dec,
       DISTANCE(POINT('ICRS', ra, dec), POINT('ICRS', 10.68458, 41.26917)) AS dist_deg
FROM basic
WHERE CONTAINS(POINT('ICRS', ra, dec), CIRCLE('ICRS', 10.68458, 41.26917, 0.1)) = 1
ORDER BY dist_deg ASC

JOINs

-- V-band magnitudes
SELECT b.main_id, b.ra, b.dec, f.flux AS Vmag
FROM basic AS b
JOIN flux AS f ON b.oid = f.oidref
WHERE f.filter = 'V' AND f.flux < 6.0
ORDER BY f.flux ASC

-- All identifiers for an object
SELECT b.main_id, i.id
FROM basic AS b
JOIN ident AS i ON b.oid = i.oidref
WHERE b.main_id = 'M  31'

-- Distance measurements
SELECT b.main_id, d.dist, d.unit, d.method
FROM basic AS b
JOIN mesDistance AS d ON b.oid = d.oidref
WHERE b.main_id = 'M  31'

Cross-Matching Identifiers Between Catalogs

SELECT b.main_id, i1.id AS hipparcos_id, i2.id AS gaia_id
FROM basic AS b
JOIN ident AS i1 ON b.oid = i1.oidref AND i1.id LIKE 'HIP %'
JOIN ident AS i2 ON b.oid = i2.oidref AND i2.id LIKE 'Gaia DR3%'
WHERE b.otype = '*' AND b.plx_value > 50

Bibliography for Objects in a Region

SELECT b.main_id, r.bibcode, r.title, r."year"
FROM basic AS b
JOIN has_ref AS hr ON b.oid = hr.oidref
JOIN ref AS r ON hr.oidbibref = r.oidbib
WHERE CONTAINS(POINT('ICRS', b.ra, b.dec), CIRCLE('ICRS', 83.633, -5.375, 0.5)) = 1
  AND r."year" >= 2020
ORDER BY r."year" DESC

Aggregation

-- Count objects by type in a region
SELECT otype, COUNT(*) AS count
FROM basic
WHERE CONTAINS(POINT('ICRS', ra, dec), CIRCLE('ICRS', 266.417, -29.008, 1.0)) = 1
GROUP BY otype
HAVING COUNT(*) > 5
ORDER BY count DESC

-- Average parallax grouped by the exact spectral-type string
SELECT sp_type, AVG(plx_value) AS mean_plx, COUNT(*) AS n
FROM basic
WHERE sp_type IS NOT NULL AND plx_value IS NOT NULL
GROUP BY sp_type
ORDER BY sp_type

Response Formats

TAP JSON

{
  "metadata": [
    {"name": "main_id", "datatype": "char"},
    {"name": "ra", "datatype": "double"},
    {"name": "dec", "datatype": "double"}
  ],
  "data": [
    ["M 31", 10.6847, 41.2687]
  ]
}

Script Response

Plain text with pipe-delimited fields. Lines starting with :: are metadata — filter them out. Check for error or not found in data lines to detect failures.

Rate Limits

No formal rate limits documented. Best practices:

  • Add time.sleep(0.5) between sequential script queries
  • Use TAP/ADQL for batch queries instead of looping over the script interface
  • Always set MAXREC or use TOP N in TAP queries to avoid accidentally downloading millions of rows
  • Very large TAP queries may time out; tighten WHERE clauses or switch to async TAP at /simbad/sim-tap/async
  • Use VOTable format for large TAP results (preserves data types and units better than JSON)

Input Sanitization

When building ADQL or script queries from user-supplied object names, sanitize inputs to prevent injection:

  • Reject control characters in object names. Use form/URL encoding for transport, and keep script commands and field names fixed
  • Escape single quotes in ADQL string literals by doubling them (' → '')
  • Limit input length (128 chars is reasonable)
  • Collapse and trim whitespace