Copies a caller-supplied byte count into a destination without inherent bounds checking.
The pack does not provide a callable signature for this record; the watched access path is shown directly.
| Arg | Access path | Purpose | Watched |
|---|---|---|---|
| Argument[0] | Argument[0] | The access path Atropos marks for this model. | ▲ sink |
Attacker's-eye view · buffer-write
Input longer than the destination overwrites adjacent memory or truncates unexpectedly.
// modeled boundary: Argument[0]
memcpy(attacker_influenced_value);
1Identify destination capacity.
2Check the write length before the call.
3Confirm truncation or rejection is explicit.
These checks require the surrounding codebase. The model names the boundary; it does not decide reachability or prove that a guard dominates every path.
Is this a bug in your code?Atropos stops here — by design
Atropos identifies Argument[0] as a buffer-write sink. It cannot see whether untrusted data reaches this call in your repository.
Lachesis is the codebase-level step: it traces reachability and guards for this symbol.
Check this symbol in Lachesis →Reviewed guidance
Safer direction: Carry destination capacity alongside the buffer and use a bounded operation with a checked length.
Unsafe shape
memcpy(dst, src, user_length);
Safer shape
if (user_length <= dst_capacity) memcpy(dst, src, user_length);
Across languages
No cross-language match.
What neutralizes it
No sanitizer of this kind is modeled.
Model facts · verbatim from the pack
| Role | Kind | Access path | Model ID | Confidence |
|---|---|---|---|---|
| sink | buffer-write | Argument[0] | c.std.memcpy.a0 | medium · corrob. 2 |
| summary | copy | Argument[1] -> Argument[0] | c.std.memcpy.a1-a0 | high · corrob. 3 |
| sink | buffer-size | Argument[2] | c.std.memcpy.a2 | high · corrob. 3 |