CVE-2026-77956 (GCVE-0-2026-77956)
Vulnerability from cvelistv5
Published
2026-08-31 00:56
Modified
2026-08-31 16:01
Severity ?
VLAI Severity ?
EPSS score ?
CWE
- CWE-94 - Improper Control of Generation of Code ('Code Injection')
Summary
Improper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.
AshAi.Actions.Prompt evaluates prompt content through EEx.eval_string/2. The documented prompt: fn input, context -> ... end form lets the prompt content be built from action arguments, so when a prompt action's text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as <%= System.cmd(...) %> therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.
This issue affects ash_ai: from 0.1.0 before 1.0.0.
References
Impacted products
| Vendor | Product | Version | |||||||
|---|---|---|---|---|---|---|---|---|---|
| ash-project | ash_ai |
Version: 0.1.0 ≤ cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:* |
|||||||
|
|||||||||
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-77956",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "total"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-08-31T16:01:20.485407Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-08-31T16:01:41.879Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/ash-project/ash_ai/security/advisories/GHSA-2g59-hg7m-qc83"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"collectionURL": "https://repo.hex.pm",
"cpes": [
"cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:*"
],
"defaultStatus": "unaffected",
"modules": [
"\u0027Elixir.AshAi.Actions.Prompt\u0027"
],
"packageName": "ash_ai",
"packageURL": "pkg:hex/ash_ai",
"product": "ash_ai",
"programFiles": [
"lib/ash_ai/actions/prompt.ex"
],
"programRoutines": [
{
"name": "\u0027Elixir.AshAi.Actions.Prompt\u0027:build_context/3"
}
],
"repo": "https://github.com/ash-project/ash_ai",
"vendor": "ash-project",
"versions": [
{
"lessThan": "1.0.0",
"status": "affected",
"version": "0.1.0",
"versionType": "semver"
}
]
},
{
"collectionURL": "https://github.com",
"cpes": [
"cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:*"
],
"defaultStatus": "unaffected",
"modules": [
"\u0027Elixir.AshAi.Actions.Prompt\u0027"
],
"packageName": "ash-project/ash_ai",
"packageURL": "pkg:github/ash-project/ash_ai",
"product": "ash_ai",
"programFiles": [
"lib/ash_ai/actions/prompt.ex"
],
"programRoutines": [
{
"name": "\u0027Elixir.AshAi.Actions.Prompt\u0027:build_context/3"
}
],
"repo": "https://github.com/ash-project/ash_ai",
"vendor": "ash-project",
"versions": [
{
"lessThan": "e9948254b5659c1143b73dc2f59f457931e64514",
"status": "affected",
"version": "4aab131d40a0bd5a8cf0b3c4eaaa59d49565f3d1",
"versionType": "git"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:*",
"versionEndExcluding": "1.0.0",
"versionStartIncluding": "0.1.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
],
"operator": "AND"
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Peter Ullrich"
},
{
"lang": "en",
"type": "reporter",
"value": "Peter Ullrich"
},
{
"lang": "en",
"type": "remediation developer",
"value": "Zach Daniel / Ash Project"
},
{
"lang": "en",
"type": "coordinator",
"value": "Jonatan M\u00e4nnchen / EEF"
}
],
"descriptions": [
{
"lang": "en",
"supportingMedia": [
{
"base64": false,
"type": "text/html",
"value": "\u003cp\u003eImproper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.\u003c/p\u003e\n\u003cp\u003e\u003ccode\u003eAshAi.Actions.Prompt\u003c/code\u003e evaluates prompt content through \u003ccode\u003eEEx.eval_string/2\u003c/code\u003e. The documented \u003ccode\u003eprompt: fn input, context -\u0026gt; ... end\u003c/code\u003e form lets the prompt content be built from action arguments, so when a \u003ccode\u003eprompt\u003c/code\u003e action\u0027s text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as \u003ccode\u003e\u0026lt;%= System.cmd(...) %\u0026gt;\u003c/code\u003e therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.\u003c/p\u003e\n\u003cp\u003eThis issue affects ash_ai: from 0.1.0 before 1.0.0.\u003c/p\u003e"
},
{
"base64": false,
"type": "text/markdown",
"value": "Improper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.\n\n`AshAi.Actions.Prompt` evaluates prompt content through `EEx.eval_string/2`. The documented `prompt: fn input, context -\u003e ... end` form lets the prompt content be built from action arguments, so when a `prompt` action\u0027s text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as `\u003c%= System.cmd(...) %\u003e` therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.\n\nThis issue affects ash_ai: from 0.1.0 before 1.0.0."
}
],
"value": "Improper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.\n\nAshAi.Actions.Prompt evaluates prompt content through EEx.eval_string/2. The documented prompt: fn input, context -\u003e ... end form lets the prompt content be built from action arguments, so when a prompt action\u0027s text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as \u003c%= System.cmd(...) %\u003e therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.\n\nThis issue affects ash_ai: from 0.1.0 before 1.0.0."
}
],
"impacts": [
{
"capecId": "CAPEC-242",
"descriptions": [
{
"lang": "en",
"value": "CAPEC-242 Code Injection"
}
]
}
],
"metrics": [
{
"cvssV4_0": {
"Automatable": "NOT_DEFINED",
"Recovery": "NOT_DEFINED",
"Safety": "NOT_DEFINED",
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "LOCAL",
"baseScore": 8.9,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"providerUrgency": "NOT_DEFINED",
"subAvailabilityImpact": "HIGH",
"subConfidentialityImpact": "HIGH",
"subIntegrityImpact": "HIGH",
"userInteraction": "NONE",
"valueDensity": "NOT_DEFINED",
"vectorString": "CVSS:4.0/AV:L/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "HIGH",
"vulnIntegrityImpact": "HIGH",
"vulnerabilityResponseEffort": "NOT_DEFINED"
},
"format": "CVSS",
"scenarios": [
{
"lang": "en",
"value": "GENERAL"
}
]
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-94",
"description": "CWE-94 Improper Control of Generation of Code (\u0027Code Injection\u0027)",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-08-31T15:08:21.204Z",
"orgId": "6b3ad84c-e1a6-4bf7-a703-f496b71e49db",
"shortName": "EEF"
},
"references": [
{
"tags": [
"vendor-advisory",
"related"
],
"url": "https://github.com/ash-project/ash_ai/security/advisories/GHSA-2g59-hg7m-qc83"
},
{
"tags": [
"related"
],
"url": "https://cna.erlef.org/cves/CVE-2026-77956.html"
},
{
"tags": [
"related"
],
"url": "https://osv.dev/vulnerability/EEF-CVE-2026-77956"
},
{
"tags": [
"patch"
],
"url": "https://github.com/ash-project/ash_ai/commit/e9948254b5659c1143b73dc2f59f457931e64514"
}
],
"source": {
"discovery": "EXTERNAL"
},
"title": "EEx template evaluation of prompt content in AshAi enables remote code execution"
}
},
"cveMetadata": {
"assignerOrgId": "6b3ad84c-e1a6-4bf7-a703-f496b71e49db",
"assignerShortName": "EEF",
"cveId": "CVE-2026-77956",
"datePublished": "2026-08-31T00:56:26.440Z",
"dateReserved": "2026-08-30T17:30:01.409Z",
"dateUpdated": "2026-08-31T16:01:41.879Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2",
"vulnerability-lookup:meta": {
"vulnrichment": {
"containers": "{\"adp\": [{\"title\": \"CISA ADP Vulnrichment\", \"metrics\": [{\"other\": {\"type\": \"ssvc\", \"content\": {\"id\": \"CVE-2026-77956\", \"role\": \"CISA Coordinator\", \"options\": [{\"Exploitation\": \"poc\"}, {\"Automatable\": \"no\"}, {\"Technical Impact\": \"total\"}], \"version\": \"2.0.3\", \"timestamp\": \"2026-08-31T16:01:20.485407Z\"}}}], \"references\": [{\"url\": \"https://github.com/ash-project/ash_ai/security/advisories/GHSA-2g59-hg7m-qc83\", \"tags\": [\"exploit\"]}], \"providerMetadata\": {\"orgId\": \"134c704f-9b21-4f2e-91b3-4a467353bcc0\", \"shortName\": \"CISA-ADP\", \"dateUpdated\": \"2026-08-31T16:01:38.306Z\"}}], \"cna\": {\"title\": \"EEx template evaluation of prompt content in AshAi enables remote code execution\", \"source\": {\"discovery\": \"EXTERNAL\"}, \"credits\": [{\"lang\": \"en\", \"type\": \"finder\", \"value\": \"Peter Ullrich\"}, {\"lang\": \"en\", \"type\": \"reporter\", \"value\": \"Peter Ullrich\"}, {\"lang\": \"en\", \"type\": \"remediation developer\", \"value\": \"Zach Daniel / Ash Project\"}, {\"lang\": \"en\", \"type\": \"coordinator\", \"value\": \"Jonatan M\\u00e4nnchen / EEF\"}], \"impacts\": [{\"capecId\": \"CAPEC-242\", \"descriptions\": [{\"lang\": \"en\", \"value\": \"CAPEC-242 Code Injection\"}]}], \"metrics\": [{\"format\": \"CVSS\", \"cvssV4_0\": {\"Safety\": \"NOT_DEFINED\", \"version\": \"4.0\", \"Recovery\": \"NOT_DEFINED\", \"baseScore\": 8.9, \"Automatable\": \"NOT_DEFINED\", \"attackVector\": \"LOCAL\", \"baseSeverity\": \"HIGH\", \"valueDensity\": \"NOT_DEFINED\", \"vectorString\": \"CVSS:4.0/AV:L/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H\", \"providerUrgency\": \"NOT_DEFINED\", \"userInteraction\": \"NONE\", \"attackComplexity\": \"LOW\", \"attackRequirements\": \"PRESENT\", \"privilegesRequired\": \"NONE\", \"subIntegrityImpact\": \"HIGH\", \"vulnIntegrityImpact\": \"HIGH\", \"subAvailabilityImpact\": \"HIGH\", \"vulnAvailabilityImpact\": \"HIGH\", \"subConfidentialityImpact\": \"HIGH\", \"vulnConfidentialityImpact\": \"HIGH\", \"vulnerabilityResponseEffort\": \"NOT_DEFINED\"}, \"scenarios\": [{\"lang\": \"en\", \"value\": \"GENERAL\"}]}], \"affected\": [{\"cpes\": [\"cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:*\"], \"repo\": \"https://github.com/ash-project/ash_ai\", \"vendor\": \"ash-project\", \"modules\": [\"\u0027Elixir.AshAi.Actions.Prompt\u0027\"], \"product\": \"ash_ai\", \"versions\": [{\"status\": \"affected\", \"version\": \"0.1.0\", \"lessThan\": \"1.0.0\", \"versionType\": \"semver\"}], \"packageURL\": \"pkg:hex/ash_ai\", \"packageName\": \"ash_ai\", \"programFiles\": [\"lib/ash_ai/actions/prompt.ex\"], \"collectionURL\": \"https://repo.hex.pm\", \"defaultStatus\": \"unaffected\", \"programRoutines\": [{\"name\": \"\u0027Elixir.AshAi.Actions.Prompt\u0027:build_context/3\"}]}, {\"cpes\": [\"cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:*\"], \"repo\": \"https://github.com/ash-project/ash_ai\", \"vendor\": \"ash-project\", \"modules\": [\"\u0027Elixir.AshAi.Actions.Prompt\u0027\"], \"product\": \"ash_ai\", \"versions\": [{\"status\": \"affected\", \"version\": \"4aab131d40a0bd5a8cf0b3c4eaaa59d49565f3d1\", \"lessThan\": \"e9948254b5659c1143b73dc2f59f457931e64514\", \"versionType\": \"git\"}], \"packageURL\": \"pkg:github/ash-project/ash_ai\", \"packageName\": \"ash-project/ash_ai\", \"programFiles\": [\"lib/ash_ai/actions/prompt.ex\"], \"collectionURL\": \"https://github.com\", \"defaultStatus\": \"unaffected\", \"programRoutines\": [{\"name\": \"\u0027Elixir.AshAi.Actions.Prompt\u0027:build_context/3\"}]}], \"references\": [{\"url\": \"https://github.com/ash-project/ash_ai/security/advisories/GHSA-2g59-hg7m-qc83\", \"tags\": [\"vendor-advisory\", \"related\"]}, {\"url\": \"https://cna.erlef.org/cves/CVE-2026-77956.html\", \"tags\": [\"related\"]}, {\"url\": \"https://osv.dev/vulnerability/EEF-CVE-2026-77956\", \"tags\": [\"related\"]}, {\"url\": \"https://github.com/ash-project/ash_ai/commit/e9948254b5659c1143b73dc2f59f457931e64514\", \"tags\": [\"patch\"]}], \"descriptions\": [{\"lang\": \"en\", \"value\": \"Improper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.\\n\\nAshAi.Actions.Prompt evaluates prompt content through EEx.eval_string/2. The documented prompt: fn input, context -\u003e ... end form lets the prompt content be built from action arguments, so when a prompt action\u0027s text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as \u003c%= System.cmd(...) %\u003e therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.\\n\\nThis issue affects ash_ai: from 0.1.0 before 1.0.0.\", \"supportingMedia\": [{\"type\": \"text/html\", \"value\": \"\u003cp\u003eImproper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.\u003c/p\u003e\\n\u003cp\u003e\u003ccode\u003eAshAi.Actions.Prompt\u003c/code\u003e evaluates prompt content through \u003ccode\u003eEEx.eval_string/2\u003c/code\u003e. The documented \u003ccode\u003eprompt: fn input, context -\u0026gt; ... end\u003c/code\u003e form lets the prompt content be built from action arguments, so when a \u003ccode\u003eprompt\u003c/code\u003e action\u0027s text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as \u003ccode\u003e\u0026lt;%= System.cmd(...) %\u0026gt;\u003c/code\u003e therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.\u003c/p\u003e\\n\u003cp\u003eThis issue affects ash_ai: from 0.1.0 before 1.0.0.\u003c/p\u003e\", \"base64\": false}, {\"type\": \"text/markdown\", \"value\": \"Improper Control of Generation of Code (Code Injection) vulnerability in ash-project ash_ai allows a remote, unauthenticated client to execute arbitrary Elixir code.\\n\\n`AshAi.Actions.Prompt` evaluates prompt content through `EEx.eval_string/2`. The documented `prompt: fn input, context -\u003e ... end` form lets the prompt content be built from action arguments, so when a `prompt` action\u0027s text incorporates request data, that attacker-controlled text is compiled and run as an EEx template (Elixir source). Content such as `\u003c%= System.cmd(...) %\u003e` therefore executes on the server before any model request is made, requiring no authentication beyond reaching a prompt action. The fix stops evaluating function-supplied prompt content as EEx; only statically configured templates are evaluated.\\n\\nThis issue affects ash_ai: from 0.1.0 before 1.0.0.\", \"base64\": false}]}], \"problemTypes\": [{\"descriptions\": [{\"lang\": \"en\", \"type\": \"CWE\", \"cweId\": \"CWE-94\", \"description\": \"CWE-94 Improper Control of Generation of Code (\u0027Code Injection\u0027)\"}]}], \"cpeApplicability\": [{\"nodes\": [{\"negate\": false, \"cpeMatch\": [{\"criteria\": \"cpe:2.3:a:ash-project:ash_ai:*:*:*:*:*:*:*:*\", \"vulnerable\": true, \"versionEndExcluding\": \"1.0.0\", \"versionStartIncluding\": \"0.1.0\"}], \"operator\": \"OR\"}], \"operator\": \"AND\"}], \"providerMetadata\": {\"orgId\": \"6b3ad84c-e1a6-4bf7-a703-f496b71e49db\", \"shortName\": \"EEF\", \"dateUpdated\": \"2026-08-31T15:08:21.204Z\"}}}",
"cveMetadata": "{\"cveId\": \"CVE-2026-77956\", \"state\": \"PUBLISHED\", \"dateUpdated\": \"2026-08-31T16:01:41.879Z\", \"dateReserved\": \"2026-08-30T17:30:01.409Z\", \"assignerOrgId\": \"6b3ad84c-e1a6-4bf7-a703-f496b71e49db\", \"datePublished\": \"2026-08-31T00:56:26.440Z\", \"assignerShortName\": \"EEF\"}",
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
}
}
Loading…
Loading…
Sightings
| Author | Source | Type | Date |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or seen somewhere by the user.
- Confirmed: The vulnerability is confirmed from an analyst perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: This vulnerability was exploited and seen by the user reporting the sighting.
- Patched: This vulnerability was successfully patched by the user reporting the sighting.
- Not exploited: This vulnerability was not exploited or seen by the user reporting the sighting.
- Not confirmed: The user expresses doubt about the veracity of the vulnerability.
- Not patched: This vulnerability was not successfully patched by the user reporting the sighting.
Loading…
Loading…