<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: AgentReady Tools</title>
    <description>The latest articles on DEV Community by AgentReady Tools (@agentreadytools).</description>
    <link>https://dev.to/agentreadytools</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4247277%2F64063a25-2fdc-480d-a495-eb995f0b2eb8.png</url>
      <title>DEV Community: AgentReady Tools</title>
      <link>https://dev.to/agentreadytools</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/agentreadytools"/>
    <language>en</language>
    <item>
      <title>A drug question your agent shouldn't answer from a web page</title>
      <dc:creator>AgentReady Tools</dc:creator>
      <pubDate>Mon, 16 Nov 2026 09:43:24 +0000</pubDate>
      <link>https://dev.to/agentreadytools/a-drug-question-your-agent-shouldnt-answer-from-a-web-page-5ej6</link>
      <guid>https://dev.to/agentreadytools/a-drug-question-your-agent-shouldnt-answer-from-a-web-page-5ej6</guid>
      <description>&lt;p&gt;This first appeared on 16 November 2026 at &lt;a href="https://agentready-tools.com/blog/drug-questions-four-official-health-apis" rel="noopener noreferrer"&gt;AgentReady Tools&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Suppose you ask an agent a narrow question about a drug. Is atorvastatin in a recruiting trial? What does PubMed return for it? Does FDA have labeling or adverse-event records? An agent that tries to answer from web pages will come back with a narrative. What you actually need is an identifier you can reuse, a status field you can check, and a timestamp for when the source last refreshed.&lt;/p&gt;

&lt;p&gt;I called four official APIs on 16 November 2026 and wrote down what they returned. All four are listed in the &lt;a href="https://agentready-tools.com" rel="noopener noreferrer"&gt;AgentReady Tools&lt;/a&gt; catalog, in the &lt;a href="https://agentready-tools.com/directory#health-biomedical" rel="noopener noreferrer"&gt;Health and biomedical data&lt;/a&gt; category. None of them charged a fee for the calls I made. They also don't share a login model, a rate limit, or a common ID. That missing common ID is the part that breaks a workflow.&lt;/p&gt;

&lt;p&gt;The data is already in JSON. The work is calling the right endpoint with an identifier that survives the next search, and not asking any source to do more than it claims to support.&lt;/p&gt;

&lt;h2&gt;
  
  
  Start with the name
&lt;/h2&gt;

&lt;p&gt;If an agent searches "Lipitor" in one system and "atorvastatin" in another, it can miss records that belong to the same drug. NLM's &lt;a href="https://agentready-tools.com/tools/rxnorm-rxnav" rel="noopener noreferrer"&gt;RxNorm / RxNav&lt;/a&gt; APIs exist to resolve that. I called &lt;code&gt;https://rxnav.nlm.nih.gov/REST/rxcui.json?name=atorvastatin&lt;/code&gt; and got a single RxCUI: &lt;code&gt;83367&lt;/code&gt;. The same host reported its dataset version as 02-Nov-2026 and &lt;code&gt;apiVersion&lt;/code&gt; 3.1.357. I treated that RxCUI as a recorded identifier for the name I looked up. I did not assume the other three APIs would accept it as a query key.&lt;/p&gt;

&lt;p&gt;NLM's &lt;a href="https://lhncbc.nlm.nih.gov/RxNav/TermsofService.html" rel="noopener noreferrer"&gt;Terms of Service&lt;/a&gt; say the RxNorm, RxTerms, Prescribable RxNorm, and RxClass APIs may be used free of charge, at no more than 20 requests per second per IP address. Over that, the servers may stop servicing requests until the rate falls, and the official &lt;a href="https://lhncbc.nlm.nih.gov/RxNav/information/FAQs.html" rel="noopener noreferrer"&gt;FAQ&lt;/a&gt; says the response is HTTP 429. A free UMLS license can raise the cap to 100 requests per second if you send your UTS API key as &lt;code&gt;Authorization: Bearer &amp;lt;key&amp;gt;&lt;/code&gt;. NLM recommends caching results for 12 to 24 hours. The APIs are updated after each RxNorm full monthly release.&lt;/p&gt;

&lt;p&gt;Most of these APIs need no license. The official &lt;a href="https://lhncbc.nlm.nih.gov/RxNav/information/FAQs.html" rel="noopener noreferrer"&gt;FAQ&lt;/a&gt; makes one exception: &lt;code&gt;/rxcui/{rxcui}/proprietary&lt;/code&gt;, which needs a UMLS license and a proxy granting ticket. RxNav's drug/drug interaction features were discontinued on 2 January 2024, so this is the wrong API for that question. The same FAQ also notes a rename to "RxNorm Tools" on 10 February 2027, with a new site URL, and says existing functionality will not be impacted.&lt;/p&gt;

&lt;p&gt;When I re-checked the Terms of Service and FAQ on 16 November 2026, both pages still carried NLM's notice that, because of a lapse in government funding, information on the website may not be up to date. The API calls above still returned HTTP 200. I can't say how that notice affects every field. An agent should surface it instead of hiding it.&lt;/p&gt;

&lt;h2&gt;
  
  
  The trial registry is a different API
&lt;/h2&gt;

&lt;p&gt;Once you have a name or a condition, &lt;a href="https://agentready-tools.com/tools/clinicaltrials-gov" rel="noopener noreferrer"&gt;ClinicalTrials.gov&lt;/a&gt; is the official NLM registry API for registered studies. The current interface is version 2, documented as OpenAPI 3.0.3 at &lt;a href="https://clinicaltrials.gov/api/oas/v2" rel="noopener noreferrer"&gt;&lt;code&gt;https://clinicaltrials.gov/api/oas/v2&lt;/code&gt;&lt;/a&gt;, with server URL &lt;code&gt;https://clinicaltrials.gov/api/v2&lt;/code&gt;. The spec I fetched lists &lt;code&gt;/studies&lt;/code&gt;, &lt;code&gt;/studies/{nctId}&lt;/code&gt;, metadata and search-area helpers, a few &lt;code&gt;/stats&lt;/code&gt; paths, and &lt;code&gt;/version&lt;/code&gt;. It declares no security schemes, and I did not send a key.&lt;/p&gt;

&lt;p&gt;A &lt;code&gt;GET&lt;/code&gt; to &lt;code&gt;/api/v2/version&lt;/code&gt; returned &lt;code&gt;apiVersion&lt;/code&gt; 2.0.5 and &lt;code&gt;dataTimestamp&lt;/code&gt; 2026-11-13T10:00:06. A &lt;code&gt;GET&lt;/code&gt; to &lt;code&gt;/api/v2/studies?pageSize=1&amp;amp;countTotal=true&lt;/code&gt; returned &lt;code&gt;totalCount&lt;/code&gt; 611771, one study record, and a &lt;code&gt;nextPageToken&lt;/code&gt;. I did not filter that studies call by drug name. The study it returned, NCT03772535, is not about atorvastatin, so that unfiltered call does not answer the opening question. Official &lt;code&gt;pageSize&lt;/code&gt; defaults to 10 and is coerced down to 1,000 if you ask for more. Pages can come back with an empty &lt;code&gt;studies&lt;/code&gt; array. If you query with only NCT IDs, filters are ignored. All of that is in the spec.&lt;/p&gt;

&lt;p&gt;I then filtered the same endpoint. &lt;code&gt;query.term=atorvastatin&lt;/code&gt;, documented in the OpenAPI as the "Other terms" query, returned &lt;code&gt;totalCount&lt;/code&gt; 1257. The first study was NCT01491958, a completed atorvastatin trial. To ask the recruiting question directly, I called &lt;code&gt;query.intr=atorvastatin&lt;/code&gt; (the official "Intervention / treatment" query) with &lt;code&gt;filter.overallStatus=RECRUITING&lt;/code&gt;, which is one of the documented status values. That call returned &lt;code&gt;totalCount&lt;/code&gt; 75. The first study was NCT07685938, "Pharmacologic Therapies to Mitigate Radiation- Associated Heart Disease," with overall status RECRUITING and an intervention named "Atorvastatin and Lisinopril." I did not review the other 74 studies. These counts will move, and a registry status is not a recommendation to use the drug.&lt;/p&gt;

&lt;p&gt;I also tried the RxCUI as a query. &lt;code&gt;query.term=83367&lt;/code&gt; and &lt;code&gt;query.intr=83367&lt;/code&gt; both returned &lt;code&gt;totalCount&lt;/code&gt; 0. I am not treating &lt;code&gt;83367&lt;/code&gt; as a ClinicalTrials.gov query key.&lt;/p&gt;

&lt;p&gt;I also tried an older info path, &lt;code&gt;/api/info/api_vrs&lt;/code&gt;, and got 404. A legacy path, &lt;code&gt;/api/legacy/full-studies&lt;/code&gt;, still returned XML, with &lt;code&gt;APIVrs&lt;/code&gt; 1.2.0 and a &lt;code&gt;DataVrs&lt;/code&gt; that matched the v2 timestamp. If your agent was written against the classic &lt;code&gt;/api/info/&lt;/code&gt; or &lt;code&gt;/api/query/&lt;/code&gt; URLs, check it. I'd point it at &lt;code&gt;/api/v2/&lt;/code&gt; now. The official OpenAPI document does not publish a request quota, and I am not going to invent one.&lt;/p&gt;

&lt;h2&gt;
  
  
  PubMed is a third system, with a tighter cap
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://agentready-tools.com/tools/pubmed-eutilities" rel="noopener noreferrer"&gt;PubMed / NCBI E-utilities&lt;/a&gt; is the official programmatic interface for PubMed and the other Entrez databases. A single unauthenticated &lt;code&gt;ESearch&lt;/code&gt; for atorvastatin (&lt;code&gt;db=pubmed&lt;/code&gt;, &lt;code&gt;retmode=json&lt;/code&gt;, &lt;code&gt;retmax=1&lt;/code&gt;) returned &lt;code&gt;count&lt;/code&gt; 13155. That number will move. The point is the shape of the response: a count, a translation of the query, and a list of PMIDs, not a page of snippets. The first PMID in that response was 43034617.&lt;/p&gt;

&lt;p&gt;I also tried &lt;code&gt;term=83367&lt;/code&gt;. That call returned &lt;code&gt;count&lt;/code&gt; 1, and the query translation was &lt;code&gt;83367[UID]&lt;/code&gt;. PubMed treated the number as a PubMed UID, not as an RxCUI. I am not using the RxCUI as a PubMed query key.&lt;/p&gt;

&lt;p&gt;NCBI's official &lt;a href="https://ncbiinsights.ncbi.nlm.nih.gov/2017/11/02/new-api-keys-for-the-e-utilities/" rel="noopener noreferrer"&gt;Insights post on E-utilities API keys&lt;/a&gt;, still on the site when I read it, is the clearest public statement of the rate limits I could fetch today. Without a key, more than three requests per second from one IP address receives an error. A free key from your NCBI account settings raises the default to 10 requests per second for that key. You pass it as &lt;code&gt;api_key&lt;/code&gt;. If several machines share one key, they share that budget.&lt;/p&gt;

&lt;p&gt;I did not treat that 2017 post as a stand-in for the full E-utilities usage-policy chapter. NCBI's Bookshelf pages presented a browser check from this network, so I could not re-read that chapter this session. If you run a heavy job, read the current policy before you loop.&lt;/p&gt;

&lt;h2&gt;
  
  
  openFDA will tell you not to trust it for care
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://agentready-tools.com/tools/openfda" rel="noopener noreferrer"&gt;openFDA&lt;/a&gt; is the FDA's Elasticsearch API for public data on drugs, devices, foods, and related records, returned as JSON. Every query goes through one endpoint for one kind of data. I called &lt;code&gt;https://api.fda.gov/drug/event.json?limit=1&lt;/code&gt; with no key and got HTTP 200. I did not filter that call by drug name. The first record was for DURAGESIC-100, not atorvastatin, so that unfiltered call does not answer the opening question either. The &lt;code&gt;meta&lt;/code&gt; object included &lt;code&gt;last_updated&lt;/code&gt; 2026-10-28, a total of 21,054,159 matching records, and this disclaimer, quoted in full:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Do not rely on openFDA to make decisions regarding medical care. While we make every effort to ensure that data is accurate, you should assume all results are unvalidated. We may limit or otherwise restrict your access to the API in line with our Terms of Service.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;A search of drug labels for &lt;code&gt;openfda.generic_name:atorvastatin&lt;/code&gt;, limited to one record, also returned 200. That response said &lt;code&gt;last_updated&lt;/code&gt; 2026-11-13, with 276 matching labels. The &lt;code&gt;meta&lt;/code&gt; object carried the same disclaimer. The label record I got listed an openFDA &lt;code&gt;rxcui&lt;/code&gt; of &lt;code&gt;259255&lt;/code&gt;, not the ingredient RxCUI &lt;code&gt;83367&lt;/code&gt; I had recorded from RxNav.&lt;/p&gt;

&lt;p&gt;A filtered adverse-event search, &lt;code&gt;patient.drug.openfda.generic_name:atorvastatin&lt;/code&gt;, also returned 200. That response said &lt;code&gt;last_updated&lt;/code&gt; 2026-10-28, with 252,398 matching records, and the same official disclaimer in &lt;code&gt;meta&lt;/code&gt;. The first record included Lipitor / atorvastatin calcium among the listed drugs.&lt;/p&gt;

&lt;p&gt;I then searched &lt;code&gt;openfda.rxcui:83367&lt;/code&gt; on both the label and event endpoints. Both returned HTTP 404 with the message "No matches found!" I am not treating the ingredient RxCUI &lt;code&gt;83367&lt;/code&gt; as an openFDA query key.&lt;/p&gt;

&lt;p&gt;The official &lt;a href="https://open.fda.gov/apis/authentication/" rel="noopener noreferrer"&gt;authentication page&lt;/a&gt; lists both of the following, on the same page: "An API key is required to make calls to the openFDA API," and standard limits for calls with no key (240 requests per minute and 1,000 per day per IP) versus calls with a free key (240 per minute and 120,000 per day per key). The key goes in the &lt;code&gt;api_key&lt;/code&gt; query parameter, or as a basic-auth username. All queries are supposed to use &lt;code&gt;https://api.fda.gov&lt;/code&gt;. Official &lt;a href="https://open.fda.gov/apis/query-parameters/" rel="noopener noreferrer"&gt;query docs&lt;/a&gt; cap &lt;code&gt;limit&lt;/code&gt; at 1,000 and &lt;code&gt;skip&lt;/code&gt; at 25,000. The about page adds that not all openFDA data has been validated for clinical or production use, and that it does not include personally identifiable patient information.&lt;/p&gt;

&lt;p&gt;If you're wiring this into an agent, the disclaimer in &lt;code&gt;meta&lt;/code&gt; should stay visible. It is part of the record.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I'd do with these four
&lt;/h2&gt;

&lt;p&gt;I'd resolve the name first in RxNorm and keep the RxCUI in the result as a recorded identifier. For ClinicalTrials.gov, PubMed, and openFDA, I searched with the generic name, not the RxCUI, because the RxCUI lookups above did not work as query keys in those systems on 16 November 2026. I'd also keep each source's own identifier: NCT ID, PMID, and whatever openFDA field you actually searched.&lt;/p&gt;

&lt;p&gt;The RxNav FAQ says a 429 means too many requests. NCBI's rate-limit post describes an error when you exceed the cap. I did not observe a 429 or a 400 in the calls above. An empty page of studies is a valid answer in the ClinicalTrials.gov spec.&lt;/p&gt;

&lt;p&gt;So, on 16 November 2026, here is what those opening questions actually returned from the official JSON, with the caveats already stated next to each call. The registry had 75 studies when I filtered intervention atorvastatin and overall status RECRUITING; the first was NCT07685938. That is a registry status, not medical advice, and I did not read the other 74. PubMed's ESearch for atorvastatin returned &lt;code&gt;count&lt;/code&gt; 13155 and a list of PMIDs. openFDA had 276 labels and 252,398 adverse-event records for &lt;code&gt;generic_name:atorvastatin&lt;/code&gt;, each with the official disclaimer.&lt;/p&gt;

&lt;p&gt;These four listings are a slice of a larger set. As of 16 November 2026 the &lt;a href="https://agentready-tools.com" rel="noopener noreferrer"&gt;AgentReady Tools&lt;/a&gt; catalog at &lt;a href="https://agentready-tools.com/tools.json" rel="noopener noreferrer"&gt;/tools.json&lt;/a&gt; has more than 250 agent-ready tools, about 20 of them in &lt;a href="https://agentready-tools.com/directory#health-biomedical" rel="noopener noreferrer"&gt;Health and biomedical data&lt;/a&gt;. I didn't re-verify the rest of that category for this post.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;For agents:&lt;/strong&gt; Machine-readable records are at &lt;a href="https://agentready-tools.com/tools/rxnorm-rxnav.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools/rxnorm-rxnav.json&lt;/a&gt;, &lt;a href="https://agentready-tools.com/tools/clinicaltrials-gov.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools/clinicaltrials-gov.json&lt;/a&gt;, &lt;a href="https://agentready-tools.com/tools/pubmed-eutilities.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools/pubmed-eutilities.json&lt;/a&gt;, and &lt;a href="https://agentready-tools.com/tools/openfda.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools/openfda.json&lt;/a&gt;, plus &lt;a href="https://agentready-tools.com/tools.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools.json&lt;/a&gt; and &lt;a href="https://agentready-tools.com/llms.txt" rel="noopener noreferrer"&gt;https://agentready-tools.com/llms.txt&lt;/a&gt;. The public catalog MCP is &lt;a href="https://agentready-tools.com/mcp" rel="noopener noreferrer"&gt;https://agentready-tools.com/mcp&lt;/a&gt; (&lt;code&gt;com.agentready-tools/catalog&lt;/code&gt;). Official vendor docs and a clinician remain the authority.&lt;/p&gt;

&lt;p&gt;None of this is medical advice, and none of these APIs asked me to treat it as such. If your agent is still answering that drug question from a pile of web pages, start with the name in RxNorm, search the other three APIs with the name unless you have confirmed how that system uses an RxCUI, and keep the disclaimers attached.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://agentready-tools.com/tools/rxnorm-rxnav" rel="noopener noreferrer"&gt;RxNorm / RxNav&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://agentready-tools.com/tools/clinicaltrials-gov" rel="noopener noreferrer"&gt;ClinicalTrials.gov API&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://agentready-tools.com/tools/pubmed-eutilities" rel="noopener noreferrer"&gt;PubMed / NCBI E-utilities&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://agentready-tools.com/tools/openfda" rel="noopener noreferrer"&gt;openFDA&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://agentready-tools.com/directory#health-biomedical" rel="noopener noreferrer"&gt;Health and biomedical data&lt;/a&gt; in the directory&lt;/li&gt;
&lt;/ul&gt;

</description>
      <category>ai</category>
      <category>api</category>
      <category>agents</category>
    </item>
    <item>
      <title>What makes a tool agent-ready: 5 checks from building a SAM.gov tool for agents</title>
      <dc:creator>AgentReady Tools</dc:creator>
      <pubDate>Sun, 15 Nov 2026 10:00:00 +0000</pubDate>
      <link>https://dev.to/agentreadytools/what-makes-a-tool-agent-ready-5-checks-from-building-a-samgov-tool-for-agents-1934</link>
      <guid>https://dev.to/agentreadytools/what-makes-a-tool-agent-ready-5-checks-from-building-a-samgov-tool-for-agents-1934</guid>
      <description>&lt;p&gt;This first appeared on 15 November 2026 at &lt;a href="https://agentready-tools.com/blog/what-makes-a-tool-agent-ready" rel="noopener noreferrer"&gt;AgentReady Tools&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;The official API for U.S. federal contract notices stores the response deadline under &lt;code&gt;reponseDeadLine&lt;/code&gt;. That's how the field appears in GSA's &lt;a href="https://open.gsa.gov/api/get-opportunities-public-api/" rel="noopener noreferrer"&gt;Get Opportunities Public API&lt;/a&gt;. If an agent looks for &lt;code&gt;responseDeadline&lt;/code&gt;, the date isn't there.&lt;/p&gt;

&lt;p&gt;A person can still get work done on SAM.gov, the official U.S. source for those notices. You click through search results, create a free account, open Account Details, and generate a public API key. You may have to re-enter a password or an emailed one-time code. An agent can't complete that part on its own. And even after an operator supplies a key, a new personal key can be tightly rate-limited.&lt;/p&gt;

&lt;p&gt;You don't always get a clean error. The agent spends its budget parsing a page, or a slightly different JSON blob every time, instead of doing the work. A shared key pasted into a chat is a bad workaround. The API's &lt;code&gt;description&lt;/code&gt; field is a URL, not the notice text, and date formats differ depending on how you fetch. This is what the official source looks like when an agent tries to use it.&lt;/p&gt;

&lt;p&gt;That was the bind we were in when we tried to get active solicitations into a shape an agent could actually use. We didn't want a shared key, and we didn't want to scrape SAM.gov pages. We also didn't want to hide how messy the official API can be.&lt;/p&gt;

&lt;p&gt;So we built &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities" rel="noopener noreferrer"&gt;SAM.gov Contract Opportunities&lt;/a&gt;, listed under the slug &lt;code&gt;sam-gov-opportunities&lt;/code&gt;. It also lives on the Apify Store. The default path streams the official public daily extract of active contract opportunities, &lt;code&gt;ContractOpportunitiesFullCSV&lt;/code&gt;, published by SAM.gov Data Services. No SAM.gov key is required. That file is up to about a day old. Each item carries &lt;code&gt;dataSource&lt;/code&gt; (&lt;code&gt;sam.gov-daily-extract&lt;/code&gt; or &lt;code&gt;sam.gov-api&lt;/code&gt;) and, in extract mode, &lt;code&gt;extractDate&lt;/code&gt; from the file's &lt;code&gt;Last-Modified&lt;/code&gt; header. Those fields are documented on the &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities" rel="noopener noreferrer"&gt;listing&lt;/a&gt; and in the &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities.json" rel="noopener noreferrer"&gt;machine-readable record&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;If you need live data, you pass your own free public API key from a SAM.gov account, and the Actor uses the official API. The key is a secret input, used only during the run, and is never written into the dataset. We also strip any &lt;code&gt;api_key&lt;/code&gt; query parameter from stored links.&lt;/p&gt;

&lt;p&gt;The tradeoff is written down. The extract needs no key and covers active notices. The API can include inactive notices in a one-year window, but a new personal key can be tightly rate-limited, and in API mode &lt;code&gt;description&lt;/code&gt; is a URL, not the notice text.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;For agents:&lt;/strong&gt; The listing slug is &lt;code&gt;sam-gov-opportunities&lt;/code&gt;. Machine-readable copies are at &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools/sam-gov-opportunities.json&lt;/a&gt;, &lt;a href="https://agentready-tools.com/tools.json" rel="noopener noreferrer"&gt;https://agentready-tools.com/tools.json&lt;/a&gt;, and &lt;a href="https://agentready-tools.com/llms.txt" rel="noopener noreferrer"&gt;https://agentready-tools.com/llms.txt&lt;/a&gt;. The public catalog MCP is &lt;a href="https://agentready-tools.com/mcp" rel="noopener noreferrer"&gt;https://agentready-tools.com/mcp&lt;/a&gt; (&lt;code&gt;com.agentready-tools/catalog&lt;/code&gt;).&lt;/p&gt;

&lt;p&gt;What we learned while building it is now five checks we use when we call a tool agent-ready. These are working notes from one build, not a standard, and not a claim that every listing in our directory already passes all five.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Can an agent call it and get a record back?
&lt;/h2&gt;

&lt;p&gt;Most of the useful web is still a page. Agents need a record: one object per thing, with field names that stay put. The misspelled deadline field is the obvious example, but the quieter problem is a key that appears in some rows and vanishes in others.&lt;/p&gt;

&lt;p&gt;Our Actor returns one normalized item per SAM.gov notice, in JSON, CSV, or Excel. Stable fields include &lt;code&gt;noticeId&lt;/code&gt;, &lt;code&gt;title&lt;/code&gt;, &lt;code&gt;solicitationNumber&lt;/code&gt;, &lt;code&gt;type&lt;/code&gt;, &lt;code&gt;postedDate&lt;/code&gt;, &lt;code&gt;responseDeadline&lt;/code&gt;, &lt;code&gt;naicsCode&lt;/code&gt;, &lt;code&gt;setAside&lt;/code&gt;, agency path, place of performance, points of contact, and the official SAM.gov link (&lt;code&gt;uiLink&lt;/code&gt;). We accept &lt;code&gt;reponseDeadLine&lt;/code&gt; and &lt;code&gt;responseDeadLine&lt;/code&gt; from the API and expose a single &lt;code&gt;responseDeadline&lt;/code&gt;. Missing values are &lt;code&gt;null&lt;/code&gt;, not an empty string.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. Is the data from an official source, and do you say how fresh it is?
&lt;/h2&gt;

&lt;p&gt;We don't scrape SAM.gov pages. That was a build decision: the default path uses the official public extract, and live mode uses the official API. Because the extract can be up to about a day old, every item carries &lt;code&gt;dataSource&lt;/code&gt;, and extract-mode items also carry &lt;code&gt;extractDate&lt;/code&gt; from the file's &lt;code&gt;Last-Modified&lt;/code&gt; header, so a caller can see which path produced the row and how fresh it is. Those fields are on the &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities" rel="noopener noreferrer"&gt;listing&lt;/a&gt; and in the &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities.json" rel="noopener noreferrer"&gt;JSON record&lt;/a&gt;. An agent will treat stale data as current if you don't say so.&lt;/p&gt;

&lt;h2&gt;
  
  
  3. Can the agent actually authenticate?
&lt;/h2&gt;

&lt;p&gt;A public API that still needs a human to click through Account Details, and maybe type a one-time code, is not ready for an agent on its own.&lt;/p&gt;

&lt;p&gt;The default extract path needs no SAM.gov key at all. We made that the default so an agent can get a record back without anyone completing the Account Details flow. The live path uses a key the operator already has, or can create on the SAM.gov Account Details page. We don't use a shared key, and we don't ask the agent to sign in to sam.gov in a browser.&lt;/p&gt;

&lt;p&gt;You still need an Apify token to run the Actor, or to reach it through Apify's MCP server. That's a platform credential, not a SAM.gov login.&lt;/p&gt;

&lt;h2&gt;
  
  
  4. Is the price visible before the call, and billed per useful result?
&lt;/h2&gt;

&lt;p&gt;A rule we now use when we review tools (this is policy, not a failed experiment from this build) is that a monthly plan with a soft cap is a poor fit for agents that retry and over-fetch. Pay-per-use only helps if the unit is something the caller can cap.&lt;/p&gt;

&lt;p&gt;On this Actor, as of 16 November 2026, the &lt;a href="https://apify.com/samgov-tools/samgov-opportunities" rel="noopener noreferrer"&gt;Apify Store listing&lt;/a&gt; prices it at $2 per 1,000 opportunities stored, or $0.002 each, on pay-per-event (event name &lt;code&gt;opportunity&lt;/code&gt;). Platform usage for the run is included. You are charged only for opportunities that are successfully stored. Notices that get filtered out, duplicates within a run, rejected records, and download or API errors don't generate a charge. The store README also documents &lt;code&gt;maxItems&lt;/code&gt;: it defaults to 10 and tops out at 10,000, and the budget is checked before every charged item. The same Actor is listed at &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities" rel="noopener noreferrer"&gt;SAM.gov Contract Opportunities&lt;/a&gt; and in the &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities.json" rel="noopener noreferrer"&gt;JSON record&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;A few limits we should say out loud: runs are independent, so the same query can charge for the same notice again, and if a run fails after some items are stored, those items may already have been billed. There is no cross-run deduplication.&lt;/p&gt;

&lt;h2&gt;
  
  
  5. Are the filters structured, and are the gaps written down?
&lt;/h2&gt;

&lt;p&gt;An agent needs inputs it can fill in, not a search box with hidden rules.&lt;/p&gt;

&lt;p&gt;You can filter by title keywords, NAICS (a prefix match on the extract, the official &lt;code&gt;ncode&lt;/code&gt; on the API), set-aside code, notice type (the official &lt;code&gt;ptype&lt;/code&gt; values, such as solicitation or sources sought), place-of-performance state, and a posted-date range of at most one calendar year. Empty input returns up to 10 recent active notices. Filters can also return nothing. That's a valid result, not an error.&lt;/p&gt;

&lt;p&gt;We also document the gaps, because an agent will hit them. The extract has no attachment list, so &lt;code&gt;resourceLinks&lt;/code&gt; is null in that mode. The Actor does not download attachments, which can have their own access rules. Keyword search is a title search, not full-text. Date formats differ between the extract and the API. And yes, we still have to accept &lt;code&gt;reponseDeadLine&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;That last one is small. It's also the kind of thing that quietly breaks a workflow if you pretend the upstream API is tidy.&lt;/p&gt;

&lt;h2&gt;
  
  
  What we're doing with this
&lt;/h2&gt;

&lt;p&gt;These five checks are how we now look at tools, including our own. &lt;a href="https://agentready-tools.com" rel="noopener noreferrer"&gt;AgentReady Tools&lt;/a&gt; is a curated directory of tools we think are closer to this bar: clean callable data, pay-per-use pricing, and some control over what an agent can do and spend. As of 16 November 2026 the catalog at &lt;a href="https://agentready-tools.com/tools.json" rel="noopener noreferrer"&gt;/tools.json&lt;/a&gt; lists more than 250 agent-ready tools, including this SAM.gov Actor and a few others we built (&lt;a href="https://agentready-tools.com/tools/url-screenshot" rel="noopener noreferrer"&gt;URL screenshot&lt;/a&gt;, &lt;a href="https://agentready-tools.com/tools/ats-jobs" rel="noopener noreferrer"&gt;ATS jobs&lt;/a&gt;). The site also publishes &lt;a href="https://agentready-tools.com/llms.txt" rel="noopener noreferrer"&gt;/llms.txt&lt;/a&gt; and the public MCP server described above.&lt;/p&gt;

&lt;p&gt;Spend Controls is a demand test, not a live product.&lt;/p&gt;

&lt;p&gt;If you're building for agents, I'd start with check 1 and check 5. Structured output and honest gaps are what we needed when the official deadline field was misspelled and a key still needed someone to click through Account Details. Write that kind of gap down first. Then look at the &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities" rel="noopener noreferrer"&gt;listing&lt;/a&gt;.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;The tool: &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities" rel="noopener noreferrer"&gt;agentready-tools.com/tools/sam-gov-opportunities&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;Machine-readable record: &lt;a href="https://agentready-tools.com/tools/sam-gov-opportunities.json" rel="noopener noreferrer"&gt;agentready-tools.com/tools/sam-gov-opportunities.json&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;Official Get Opportunities API: &lt;a href="https://open.gsa.gov/api/get-opportunities-public-api/" rel="noopener noreferrer"&gt;open.gsa.gov/api/get-opportunities-public-api&lt;/a&gt;
&lt;/li&gt;
&lt;/ul&gt;

</description>
      <category>ai</category>
      <category>mcp</category>
      <category>api</category>
      <category>agents</category>
    </item>
  </channel>
</rss>
